Exponential and logarithmic functions for floating-point numbers

Abstract

The present disclosure relates to exponential and logarithmic function approximations for floating point numbers. The methods presented herein enable efficient approximation of exponential or logarithmic functions using values in half-precision floating point format. In at least one embodiment, an activation function of a neural network can perform tasks such as converting raw scores from a network into probabilities using an exponentiation of each output. A fixed point representation of each raw value can be generated that includes a bias factor and used to generate an intermediate representation having a determined number of bits. The intermediate representation can be incremented such that the mantissa corresponds to the fixed point representation. The decimal point can be shifted to obtain a value that is an approximation of the exponent of the raw value and can be used to determine a probability (or other such value) for a corresponding computational operation. At least some of these operations can be reversed to similarly obtain an approximation of a logarithmic function.

Description

Background Technology

[0001] Using exponential or logarithmic functions can be beneficial in various situations. These include, for example, determining whether parameter values ​​grow or decay over time, or using them in normalized exponential (SoftMax) activation layers of neural networks to determine the probability of inference values ​​or classifications, and so on. Unfortunately, calculating exponential values ​​can be very computationally expensive, especially in operations involving the calculation and summation of multiple exponential values. Previous methods attempted to reduce computational costs by approximating exponential functions, but such approximations typically require conversion to and from integer values. While such integer-based operations are relatively efficient, they are often implemented for specific formats, such as single-precision formats. For other formats, methods requiring integer conversions may be inefficient in performance, or in some cases, may not be feasible with certain hardware. As an example, performing such conversions on half-precision formats is difficult on some graphics processing units (GPUs). Several methods have been used to attempt to improve performance when using exponential functions, including direct numerical methods (e.g., interpolation polynomials, rational approximations, and iterative techniques that utilize various algebraic properties of exponential and logarithmic functions). Furthermore, the efficiency of the IEEE floating-point representation and the hardware performing these operations (including bit manipulations) can be utilized. These methods have not yet proven to be successful alternatives or implementation options for exponential functions, at least on certain types of hardware. Attached Figure Description

[0002] Various embodiments according to this disclosure will now be described with reference to the accompanying drawings, in which: Figure 1 Components of an example system capable of inferring the classification of input data according to at least one embodiment are shown; Figure 2A An example architecture of a classification network according to at least one embodiment is shown; Figure 2B An example SoftMax activation function that can be used in at least one embodiment is shown; Figure 3A and Figure 3B The illustration shows an example method for approximate exponential operation of raw values ​​inferred by a machine learning network according to at least one embodiment; Figure 4A A relative error diagram of a half-precision approximation according to at least one embodiment is shown; Figure 4B An exemplary half-precision floating-point format according to at least one embodiment is shown; Figure 5A An example process for reasoning about one or more classifications according to at least one embodiment is shown; Figure 5B An example process for approximating an exponential function according to at least one embodiment is shown; Figure 6 Components of a distributed system according to at least one embodiment are shown, the distributed system being used to generate and provide image content using object and scene models (including textures); Figure 7A The inference and / or training logic according to at least one embodiment is illustrated; Figure 7B The inference and / or training logic according to at least one embodiment is illustrated; Figure 8 An example data center system according to at least one embodiment is shown; Figure 9 A computer system according to at least one embodiment is shown; Figure 10 A computer system according to at least one embodiment is shown; Figure 11 At least a portion of a graphics processor according to one or more embodiments is shown; Figure 12 At least a portion of a graphics processor according to one or more embodiments is shown; Figure 13 This is an example data flow diagram of an advanced computing pipeline according to at least one embodiment; Figure 14 This is a system diagram of an example system for training, adapting, instantiating, and deploying machine learning models in an advanced computing pipeline, according to at least one embodiment; and Figure 15A and Figure 15B A data flow diagram of a process for training a machine learning model according to at least one embodiment is shown, as well as a client-server architecture for enhancing annotation tools using a pre-trained annotation model. Detailed Implementation

[0003] Various embodiments will be described in the following description. Specific configurations and details are set forth for illustrative purposes in order to provide a full understanding of the embodiments. However, it will be apparent to those skilled in the art that these embodiments can be practiced without specific details. Furthermore, well-known features may be omitted or simplified so as not to obscure the embodiments being described.

[0004] The systems and methods described herein can be used by, but are not limited to, non-autonomous vehicles or machines, semi-autonomous or autonomous vehicles or machines (e.g., in one or more advanced driver assistance systems (ADAS), one or more in-vehicle infotainment systems, one or more emergency vehicle detection systems), manned and unmanned robots or robotic platforms, warehouse vehicles, off-road vehicles, vehicles connected to one or more trailers, spacecraft, ships, shuttles, emergency response vehicles, motorcycles, electric or motorized bicycles, aircraft, engineering vehicles, trains, submarines, remotely operated vehicles (such as drones), and / or other vehicle types. Furthermore, the systems and methods described herein can be used for a variety of purposes, such as, but not limited to, machine control, machine motion, machine actuation, synthetic data generation, generative AI, model training or updating, perception, augmented reality, virtual reality, mixed reality, robotics, security and surveillance, simulation and digital twins, autonomous or semi-autonomous machine applications, deep learning, environmental simulation, data center processing, conversational AI, optical transport simulation (e.g., ray tracing, path tracing, etc.), collaborative content creation for 3D assets, generative AI, cloud computing, and / or any other suitable application.

[0005] The disclosed embodiments may be included in a variety of different systems, such as automotive systems (e.g., in-vehicle infotainment systems for autonomous or semi-autonomous machines, perception systems for autonomous or semi-autonomous machines), systems implemented using robots, aerial systems, medical systems, rowing systems, intelligent area monitoring systems, systems for performing deep learning operations, systems for performing simulation operations, systems for performing digital twin operations, and systems implemented using edge devices, systems containing one or more virtual machines (VMs), systems for performing synthetic data generation operations, systems implemented at least partially in a data center, systems for performing conversational AI operations, systems implementing one or more language models such as Large Language Models (LLM), Visual Language Models (VLM), Multimodal Language Models, etc., systems for performing generative AI operations (e.g., using one or more language models, converter models, etc.), systems for performing optical transmission simulations, systems for performing collaborative content creation of 3D assets, systems implemented at least partially using cloud computing resources, and / or other types of systems.

[0006] One or more parameters of a content-generating environment can be generated using methods according to various embodiments. In at least one embodiment, parameters of the content-generating environment can be generated using a trained machine learning (ML) and / or artificial intelligence (AI) system (e.g., a large language model (LLM) or a visual language model (VLM)), such as, but not limited to, camera settings for displaying objects in a scene, scene lighting, video parameters, etc. These parameters can be based on input provided by a user or user agent to a trained language model (e.g., an LLM, VLM, etc.), which can then generate one or more settings based on the input. Various embodiments can be used to generate settings in two-dimensional (2D) or three-dimensional (3D) settings. For embodiments involving one or more language models (i.e., one or more LLMs, one or more VLMs, or a combination of LLMs and VLMs), one or more language models can receive input (e.g., prompts, requests, queries, etc.) that is parsed or otherwise formatted to generate deterministic output. For example, input provided to a language model may include a specific format of the output, an example of the expected output, a specific list of parameters and their respective formatting, etc. An input generator (e.g., a cue generator) can be driven or otherwise guided by one or more AI and / or ML systems to generate this input based on initial input received from a user, device, agent, etc. The modified input generated by the input generator can then be provided to a language model, which will generate a set of output parameters. This output can be further evaluated by a reviewer or other system to ensure its appropriateness. Subsequently, configuration files can be generated and / or the parameters can be provided directly to the environment to configure different components (e.g., camera settings, lighting, etc.) based on the parameters generated by the language model.

[0007] In some examples, the machine learning models described herein (e.g., deep neural networks, language models, LLMs, VLMs, multimodal language models, perceptual models, tracking models, fusion models, transformer models, diffusion models, encoder-only models, decoder-only models, encoder-decoder models, neural rendering field (NERF) models, etc.) can be packaged into microservices—such as inference microservices (e.g., NVIDIA NIM)—which can contain containers (e.g., operating system (OS) level virtualization packages) that can contain an application programming interface (API) layer, a server layer, a runtime layer, and / or at least one model “engine.” For example, an inference microservice can contain the container itself as well as one or more models (e.g., weights and biases). In some cases, such as when one or more machine learning models are small enough (e.g., have a small enough number of parameters), one or more models can be contained within a container. In other examples (e.g., when the models are large), the models can be hosted / stored in the cloud (e.g., in a data center), and / or hosted locally and / or at the edge (e.g., on a local server or computing device, but outside the container). In these embodiments, the model can be accessed via one or more APIs (e.g., REST APIs). Therefore, in some embodiments, the machine learning models described herein can be deployed as inference microservices to accelerate model deployment on any cloud, data center, or edge computing system while ensuring data security. For example, an inference microservice may include one or more APIs, pre-configured containers (for simplified deployment), an optimized inference engine (e.g., execution software built using standardized AI model deployments, such as NVIDIA's Triton Inference Server, and / or one or more APIs for high-performance deep learning inference, which may include inference runtime and model optimizations for providing low latency and high throughput for production applications, such as NVIDIA's TensorRT), and / or enterprise management data for telemetry (e.g., including identity, metrics, health checks, and / or monitoring).

[0008] One or more machine learning models described herein can be included as part of a microservice along with acceleration infrastructure that can be deployed with a single command and / or orchestrated and automatically scaled on the acceleration infrastructure using a container orchestration system (e.g., from a single device to data center scale). Therefore, an inference microservice may include one or more machine learning models (e.g., models optimized for high-performance inference), inference runtime software for executing one or more machine learning models and providing output / response to inputs (e.g., user queries, prompts, etc.), and enterprise management software for providing health checks, authentication, and / or other monitoring. In some embodiments, the inference microservice may include software for in-situ replacement and / or updating of the machine learning models. Upon replacement or update, the software performing the replacement / update may maintain the user configurations of the inference runtime software and the enterprise management software.

[0009] Methods according to various exemplary embodiments can provide efficient approximations of various functions, such as exponential and / or logarithmic functions, which may be useful for operations performed in neural network layers. In at least one embodiment, the SoftMax function can be used to perform tasks such as converting the network's raw output scores into probabilities using the exponent of each output score. The output can be received as a half-precision number or a floating-point number, which may not be convertible to higher precision, partly due to register limitations or other such factors. Methods according to at least one embodiment can manipulate the bits of a number to perform integer exponentiation using IEEE floating-point representation and linear interpolation by pushing bits into the mantissa. In one example, the output value can be received in floating-point format, and a fixed-point representation of that value, including a bias factor, can be generated. An intermediate representation of the value, having a defined number of bits, can be generated, and a value that can be added to 1 can be added to the intermediate representation to increase the mantissa representation of the fixed-point representation. The decimal point of the value can then be shifted by the appropriate number of bits to obtain a value that is an exponential approximation of the original output value. Afterward, any sign bits that may have been left over from the shift operation can be cleared. The floating-point precision value can then be used, along with the exponent value, to determine the probability (or other similar value) of the corresponding computational operation. At least some of these operations can be reversed to obtain a similar approximation of the logarithmic function using floating-point precision values.

[0010] Based on the teachings and suggestions contained herein, those skilled in the art will understand that variations of this function and other similar functions may also be used within the scope of various embodiments.

[0011] Figure 1An example system 100 according to at least one embodiment is illustrated, which can be used to train and use one or more machine learning models to perform inference. In this example, the machine learning model may be a classification model trained to infer one or more classifications and their associated probabilities for various input data instances. An untrained classifier model 104 (e.g., a convolutional neural network (CNN) or other similar model) may be provided for training using a model training system 106. The model training system 106 may train the untrained classifier model 104 using a set of labeled and / or unlabeled classification data 102 (depending in part on the type of training performed) to produce a trained classifier 114 that is capable of inferring the classification of unidentified objects at least within the range of classifications represented in the classification data 102. One or more instances of the trained classifier 114 may then be used to perform inference operations, for example, by hosting them on a server 108 or other computing resources in a data center or other computing environment. In one example use case, a user can submit a user query using client device 118 (e.g., a desktop or laptop computer). This query is directed to server 108 for classification performed by a trained classifier 114. The query can contain data on one or more objects to be classified, such as words in a text query or objects represented in an image query. The trained classifier 114 can process the query and output one or more classifications, each with a corresponding probability value representing the degree of certainty that the classification is correct. In this example, the trained classifier 114 includes a SoftMax activation layer as the final layer, which receives raw values ​​generated by the intermediate layers of the trained classifier and converts these values ​​into probabilities. The trained classifier 114 can output inferred classifications 120 and their probability values, which can be returned to client device 118 in response to a user query, stored in classification data repository 102 for further training of the classifier model, and / or directed to other suitable recipients.

[0012] However, as mentioned above, operations performed in layers such as the SoftMax activation layer can be computationally intensive. Server 108 may have certain processing hardware, such as at least one central processing unit (CPU) 110 and one or more graphics processing units (GPUs) 112, as well as data processing units (DPUs) or other similar options. For example, in examples requiring the use of one or more GPUs for inference operations, the capacity of individual GPUs may be limited, or certain types of operations may be more expensive to perform on these GPUs. For instance, generating probabilities in the SoftMax layer requires computing an exponential function, which can be very computationally intensive. Various approximations can be made to the exponential function (or logarithmic function, etc.), but this typically requires converting floating-point values ​​to integer values, which is computationally intensive and relatively slow, unintentionally increasing inference latency. If an attempt is made to avoid using the exponential function, the model needs to be retrained or a different model needs to be trained to implement this alternative. Even using half-precision values ​​instead of full-precision values, the overhead of some operations can still be very high, at least for some hardware.

[0013] Methods according to various embodiments can perform approximations of exponential, logarithmic, and other such functions in floating-point format. These values ​​can be received in half-precision and maintained throughout the computation without conversion to single-precision or integer values. This usage is beneficial for any use case or operation requiring efficient execution of exponential or logarithmic functions, such as various inference and training operations performed using machine learning models.

[0014] Figure 2A An example of a classification module architecture 200 that can be used according to at least one embodiment is illustrated. In this example, the model contains multiple node layers, including an input layer 202, multiple intermediate layers 204, and an output layer 206 (also referred to as the final layer or SoftMax layer, where the final layer is used with the SoftMax activation function). In such a model, nodes in adjacent layers will typically be fully connected. It should be understood that a real model will have more intermediate layers, each containing more nodes, each node representing a computation to be performed. In one example, the SoftMax output layer might receive 64 raw values ​​from the nodes of the last intermediate layer 204 and need to perform an exponential operation on each of these 64 values ​​and sum them as part of the SoftMax activation operation to generate one or more corresponding probability values. Figure 2BFigure 250 illustrates an example approximation of the SoftMax activation function according to at least one embodiment. The SoftMax function is a normalization exponential function that can be used to transform a set (or vector, etc.) or real numbers into a probability distribution of several potential outcomes. In a classifier network, the SoftMax function is used to normalize the network's output to a probability distribution of several predicted output classifications.

[0015] Figure 3A The illustration shows an example of the SoftMax function 308, including a view of the values ​​of node 302 (including bias node 304) in an example neural network, which can be used when performing calculations with this function. As previously mentioned, the SoftMax function 308 is typically used in the last layer of a neural network, such as one trained to generate classification inference. The SoftMax function 308 can be used to convert scores generated by such networks into a set of probabilities. Figure 3B As shown, the final value of x can be determined using individual values ​​352 of x from the final network layer, where each... x The value 352 is the value of the previous level node. z The value 356 is a function of the corresponding network weight 354 of the corresponding node. The transformation from the original value to a probability can be performed in part by taking the exponent of each output value and normalizing these values. Normalization can be achieved, for example, by dividing these individual exponent values ​​by the sum of all exponent values, such as... Figure 3A Example mathematical expression 308 is shown in the example. This normalization helps ensure that the output values ​​are in the range of 0 to 1 and that the sum is 1, thus effectively corresponding the output values ​​to probabilities. In example mathematical expression 308 of the SoftMax function, x i Represents a given class i The input to the SoftMax function. As mentioned earlier, the denominator of this mathematical expression 308 represents the sum of the exponents of the scores of all classes in the output layer 306 of the network.

[0016] For example, the network's task might be to classify a group of images. The last layer of the network might output a vector containing values ​​corresponding to each class. These values ​​can be converted into a probability distribution for each class using the SoftMax function. The SoftMax function accepts a vector of real numbers (or a column matrix, etc.). x As input, this vector x This represents the output from the last layer (306) of the neural network. (Vector) x Each element in the array can be exponentially operated on to ensure that all values ​​are positive. Figure 3AThe numerator of the mathematical expression 308 shown represents the exponential operation. Then, the value of the exponential operation can be divided by the sum of the values ​​of all exponential operations, thus ensuring that the sum of the generated probability distributions is 1.0, as previously described.

[0017] While other functions can be used, the SoftMax function offers several advantages, at least in certain operations or use cases. One advantage is the normalization of probability values, ensuring that all output values ​​lie between 0 and 1. As mentioned earlier, the sum of all outputs of the SoftMax function is always equal to 1. The SoftMax function can also transform the raw output into probabilities, making the predictions of classification (or other similar) networks potentially easier to understand and analyze.

[0018] However, as mentioned earlier, this approach may encounter performance issues when implemented on certain hardware. For example, calculating the SoftMax function on the values ​​of the activation layer can lead to performance problems, at least in part due to the computational cost of calculating the exponent values ​​themselves. As mentioned earlier, for values... x It is necessary to calculate the exponential function, which has the following properties: e The exponent is a number with the base of Euler's number 2.718. Methods according to various embodiments can effectively approximate the exponential function, thereby reducing the cost of computational functions (such as the SoftMax function).

[0019] In at least one embodiment, an approximation of a floating-point number can be provided using exponential and / or logarithmic functions. This may involve manipulating the bits of the logarithm to obtain an integer exponentiation operation. This can be done by using the IEEE representation of the floating-point number, where linear interpolation is performed by overflowing bits into the mantissa, or by multiplying the portion of the floating-point number representing the significant digits of the number by the exponent of the base to obtain the actual value of the number. In this example, given a half-precision value... x The approximation needs to be performed exponentially within the given calculation. A first step in the approximation could be adding an exponential bias. Alternatively, a bias constant can be subtracted from the value to improve the accuracy of the approximation. The chosen constant can be a value determined to minimize the maximum relative error, for example, given by: In this example method, select an integer. f This is to ensure the value is in fixed-point format. The length of a fixed-point number depends in part on the number of bits available in the specific implementation. For half-precision, this can correspond to attempting to obtain the largest possible range of results. f=5. The denominator uses log2 to convert this value to a natural base. In this example, shifting left by 5 bits results in a 5-digit integer value, followed by a decimal point, and then 5 decimal places. This is appropriate because for half-precision representation, the exponent is defined as having 5 digits, while the entire mantissa has 10 digits. At this stage, the middle representation y has not yet undergone exponentiation and is a fixed-point number with zeros to the left of the decimal point and all digits to the right.

[0020] The next step in this example process is to add the value 1 to y. After adding this value, the mantissa of the number will contain... y The numerator is a fixed-point representation with an exponent of 1. A shift can then be performed, for example, by the number of bits in the exponent (e.g., half-precision 5). However, the sign bit, which may be what remains after the shift, can be cleared. In this example embodiment, the half-precision result can be given by the following formula: And returns -ashalf(asint16(y)<<5). For all exponential results that can be stored in half-precision x values, the relative error will be approximately in the range of [-0.04, 0.04], such as... Figure 4A As shown, its supported interval is [-2, 2]. A logarithmic implementation can perform these operations in reverse, where this method uses an "OR" mask of floating-point values ​​of 1 instead of a bitmask. The relative error between this approximation and the true value of the half-precision exponent can reach approximately 4%. It is worth noting that at half precision, using a more accurate approximation can guarantee an achievable precision of approximately 2%.

[0021] In addition to bit manipulation, this method can also utilize floating-point representations, such as IEEE floating-point representation. This avoids the need for conversions between integer values ​​and avoids performing integer operations on 16 bits as in previous methods, which can significantly reduce the computational and resource costs of these operations. Performing these operations in half-precision can also significantly reduce the computation time per operation. Furthermore, using floating-point values ​​and some minor bit operations to perform these operations allows them to be performed in a wide variety of formats. In at least some embodiments, half-precision or floating-point numbers received as input cannot be converted to higher-precision values ​​due to memory or storage register limitations. The estimation of exponent values ​​(and the sum of exponent values) can be performed for any half-precision input or input using floating-point values.

[0022] Figure 4B The illustration shows an IEEE 754 half-precision floating-point format 450 that can be used according to at least one embodiment. This format 450 contains a 16-bit word, which is divided into a 1-bit sign indicator. s 452 and 5-bit bias index e454, and 10 decimal places. f 456. As mentioned earlier, the IEEE representation of a floating-point number includes the exponent and mantissa, and can be given, for example, by the following formula: 6.6 10 x 10 -34 The first part corresponds to the mantissa, which defines the non-zero part of the number. The second part corresponds to the exponent, which defines the number of positions the decimal point will be moved through. The exponent is positive when the decimal point moves to the right and negative when it moves to the left. Because half-precision numbers use 16 bits for storage in this method, they typically require less memory than single-precision numbers (using 32 bits) or double-precision numbers (using 64 bits). However, because these numbers use fewer bits for storage, half-precision numbers have lower precision than single-precision or double-precision numbers.

[0023] Floating-point representation is advantageous when algorithms involve large or unknown dynamic ranges, such as integrators in feedback loops, or when the algorithm uses operations that are difficult to design using fixed-point representation. As mentioned earlier, half-precision data types occupy only 16 bits of memory, but their floating-point representation allows them to handle a wider dynamic range than integer or fixed-point data types of the same size. This makes half-precision particularly suitable for certain image processing and graphics applications. For example, when combined with deep neural networks, the time required for training and inference can be significantly reduced. The memory footprint of lookup tables can be further reduced by using half-precision as the storage format for lookup tables. Approximate calculations can avoid conversion to integers, directly converting to half-precision floating-point format and then performing bit manipulations, which can be done very quickly, resulting in faster and more lightweight computations.

[0024] In another example, x It can represent a half-precision input value that requires exponentiation. For example, if the input value is 100, it can be represented as... It indicates, but It cannot be represented using half precision because it would cause a register overflow and, at least in some hardware, would be represented as an "illegal" number. In the above equation... y It is a fixed-point representation of the value x, with bias applied. It can be derived... y The value is used to obtain an intermediate representation before exponentiation is performed. y The middle representation is a fixed-point number with zeros to the left of the decimal point and all significant digits to the right. Adding the value 1 to this middle representation and then shifting those digits to the left yields the exponent of the value. If... x In this example, if it is not an integer, then... xThis is converted to a fixed-point value that can be generalized as an integer. For a fixed-point value, the format is zero followed by a decimal point and a tens digit, whereas before division, it is simply a tens digit. This method avoids converting x to an integer representation when x is input as a floating-point number and needs to be converted to a fixed-point number. As mentioned earlier, manual conversion can be performed, but in at least some embodiments, the range of conversions is limited, partly due to limitations in the hardware used to perform the operation. An approximation method can be used instead, where a value with a bias factor is generated. x The mantissa is a fixed-point representation, and this representation is used to generate an intermediate representation with a defined number of bits. The value 1 can then be added to this intermediate representation to make the mantissa represent the fixed-point representation (values ​​from 0 to 1 bits). The format "1.bits" forces the mantissa to have exactly the required bits. In at least some hardware, only formats like "1.bits" can match floating-point values ​​to fixed-point values. This method can be used with hardware that employs floating-point normalization, where bits can be shifted as needed because the value itself is correctly positioned.

[0025] Figure 5A An example process 500 according to at least one embodiment is illustrated, which can be performed to generate an inferential classification of one or more objects or elements represented in input data. It should be understood that, for these and other processes described herein, unless explicitly stated otherwise, additional, fewer, or alternative steps may exist within the scope of various embodiments, performed in a similar or alternative order, or at least partially in parallel. Furthermore, although this example will be discussed with regard to classification, other types of inference may also be performed within the scope of various embodiments, and these inferences may also benefit from exponential, logarithmic, or other such approximations. This is particularly applicable when it is advantageous to perform fast exponential operations, such as in computations like SoftMax activation determination. In this example process 500, input data is provided 502 to a classification neural network. The input data may contain representations of one or more objects or elements to be classified. Various operations 504 can be performed on the input data at various layers of the network, each layer comprising several hidden or intermediate layers with network node weights determined during training. A set of raw output values ​​may be received 506 to the final output layer of the nodes, such as the SoftMax activation layer of the classification network. In this layer, exponential approximations of these original values ​​can be generated using half-precision values ​​received from intermediate layers. The classification neural network can then generate probabilities for one or more categories as output values, these probabilities being determined using exponential approximations of half-precision floating-point values. One or more categories and probabilities can be provided in response to the provided input data, or otherwise stored or transmitted to the target receiver or destination.

[0026] Figure 5BThe illustration depicts an example process 550, which can be executed according to at least one embodiment, to generate an exponential approximation. While the discussion here pertains to its use in activation layers of a classifier network (e.g., regarding...),... Figure 5A The process described in [the original text] is not applicable to all machine learning (or other) operations, but such an approximation can be used for any machine learning (or other) operation where efficient exponentiation or other such computations need to be performed on one or more input values. In this example, receiving 552 (or otherwise obtaining or generating) will result in an exponentially calculated value in floating-point format. A fixed-point representation of this value, 554, can be generated, containing an exponent bias factor. Using this fixed-point representation, an intermediate representation of the value, 556, can be generated, having a target or defined number of bits (e.g., 10 or 16 bits). The intermediate representation can be incremented (i.e., it can have an added value of 1) so that the mantissa of the intermediate representation corresponds to the fixed-point representation. The decimal point of the incremented intermediate representation can be shifted by 560 (e.g., right-shifted by five placeholders), thus producing an exponential approximation of the received value. This exponential approximation can then be used for specific operations, such as determining and / or providing probabilities for one or more inferences generated by a trained neural network. This approach is applicable not only to linear approximations but also to higher-order approximations.

[0027] This method allows for exponential, logarithmic, or other such approximations using half-precision floating-point values. For values ​​with 10 mantissas, proper placement of the decimal point allows the 5 bits to the left of the decimal point to represent the exponent and the 5 bits to the right of the decimal points to represent the fractional part, thus enabling its use with the IEEE floating-point representation of the value. This method allows for linear interpolation between two exponent values. As mentioned earlier, adding a value of 1 allows the mantissa to be represented in a fixed-point representation, which can then be shifted as needed. A match exists between fixed-point and floating-point values ​​only when the value before the decimal point is 1 and the value after the decimal point is numeric. Because this operation involves a relatively lightweight "fused multiply-add" operation followed by bit shifting, it is relatively fast. The quality of the approximation is also relatively high, as the result is within approximately 4% of the half-precision result without this approximation. As mentioned earlier, performing these steps in reverse order also produces a logarithmic approximation.

[0028] The various methods proposed in this paper are lightweight enough to be executed near real-time in various locations, such as client devices or cloud servers. Such processing can be performed on content generated or received by a client device or server, or content received from an external source, such as content received from a cloud server 620 or a third-party service 660 via at least one network. In some cases, the processing, generation, synthesis, and / or determination of at least a portion of such content can be performed by one of these other devices, systems, or entities and then provided to client device 602 (or other such recipients) for presentation or other such purposes.

[0029] For example, Figure 6An example network configuration 600 is illustrated, which can be used to provide, generate, modify, encode, process, and / or transmit various types of data or content. In at least one embodiment, client device 602 can use components of content manager 604 and / or content application 612 on client device 602, as well as data locally stored on the client device, to generate or receive session data. In at least one embodiment, content application 624 running on server 620 (e.g., cloud server or edge server) can initiate a conversation associated with at least one client device 602, which can utilize a conversation manager and user data stored in user database 636, and can enable content manager 626 to determine one or more digital assets (e.g., implicit and / or explicit object representations) from asset library 634. Content manager 626 can work in conjunction with at least one machine learning module 628 to generate inference results that may be based in part on assets in asset library 634. In other embodiments, machine learning model 630 can be trained by training module 628 using assets in asset library 634 and other potential training data sources. In at least one embodiment, one or more machine learning models 630 may be trained to perform exponential or logarithmic approximations during inference. In at least one embodiment, the content application 624 may work with one or more encoders, transcoders, and / or compressors that can perform tasks such as encoding, decoding, compressing, and / or decompressing textures, images, or other such resources or content instances, where different compression or encoding methods may be suitable for different operations, such as storage and processing. At least a portion of the content may be transmitted to the client device 602 using a suitable transmission manager 622 for delivery via download, streaming media, or other transmission channels. An encoder may be used to encode and / or compress at least a portion of the data before transmitting it to the client device 602. In at least one embodiment, a client device 602 receiving such content may provide this content to a corresponding content application 604, which may also or optionally include a graphical user interface 610, a content application 612, and a machine learning module 614 for reasoning, providing, synthesizing, rendering, mixing, modifying, or using the content on or through the client device 602 for presentation (or other purposes). A decoder may also be used to decode data received through one or more networks 640 for presentation via the client device 602, such as presenting image or video content via a display 606, and presenting audio (e.g., sound and music) via at least one audio playback device 608 (e.g., a speaker or headphones).In at least one embodiment, at least a portion of the content may already be stored on, rendered on, or accessible to client device 602, so at least that portion of the content does not need to be transmitted over network 640, for example, the content may have been previously downloaded or stored locally on a hard disk or optical disc. In at least one embodiment, the content may be transmitted from server 620 or user database 636 to client device 602 using a transmission mechanism such as data streaming. In at least one embodiment, at least a portion of the content may be obtained, enhanced, and / or streamed from other sources (e.g., third-party service 660 or other client device 650), which may also include an application 662 for generating, enhancing, or providing the content. In at least one embodiment, a portion of the functionality may be executed using multiple computing devices or multiple processors in one or more computing devices, for example, a combination of CPU and GPU.

[0030] In this example, these client devices can include any suitable computing device, such as desktop computers, laptops, set-top boxes, streaming media devices, game consoles, smartphones, tablets, VR headsets, AR goggles, wearable computers, or smart TVs. Each client device can submit requests across at least one wired or wireless network, which can include the Internet, Ethernet, a local area network (LAN), or a cellular network, and other such options. In such an example, these requests can be submitted to an address associated with a cloud provider that can operate or control one or more electronic resources within the cloud provider's environment, such as data centers or server farms. In at least one embodiment, such requests can be received or processed by at least one edge server located at the network edge and outside at least one security layer associated with the cloud provider's environment. In this way, latency can be reduced by enabling client devices to interact with servers in closer proximity, while also improving the security of resources within the cloud provider's environment.

[0031] In at least one embodiment, such a system can be used to perform graphics rendering operations. In other embodiments, such a system can be used for other purposes, such as providing image or video content to test or validate autonomous machine applications, or for performing deep learning operations. In at least one embodiment, such a system can be implemented using side devices, or may include one or more virtual machines (VMs). In at least one embodiment, such a system can be implemented at least partially in a data center or at least partially using cloud computing resources.

[0032] Reasoning and training logic Figure 7AInference and / or training logic 715 is shown for performing inference and / or training operations associated with one or more embodiments. The following is in conjunction with... Figure 7A and / or Figure 7B Provide details about reasoning and / or training logic 715.

[0033] In at least one embodiment, inference and / or training logic 715 may include, but is not limited to, code and / or data storage 701 for storing forward and / or output weights and / or input / output data, and / or other parameters configuring neurons or layers of a neural network trained for and / or used for inference in one or more embodiments. In at least one embodiment, training logic 715 may include or be coupled to code and / or data storage 701 for storing graph code or other software to control timing and / or sequence, wherein weight and / or other parameter information is loaded to configure logic, including integer and / or floating-point units (collectively, arithmetic logic units (ALUs)). In at least one embodiment, code (such as graph code) loads weight or other parameter information into the processor ALU based on the architecture of the neural network to which the code corresponds. In at least one embodiment, code and / or data storage 701 stores weight parameters and / or input / output data of each layer of a neural network trained or used in one or more embodiments during forward propagation of input / output data and / or weight parameters during training and / or inference using one or more embodiments. In at least one embodiment, any portion of the code and / or data storage 701 may be included within other on-chip or off-chip data storage, including the processor's L1, L2, or L3 cache or system memory.

[0034] In at least one embodiment, any portion of the code and / or data storage 701 may be internal or external to one or more processors or other hardware logic devices or circuits. In at least one embodiment, the code and / or data storage 701 may be a cache memory, dynamic random-addressable memory (“DRAM”), static random-addressable memory (“SRAM”), non-volatile memory (e.g., flash memory), or other storage. In at least one embodiment, the choice of whether the code and / or data storage 701 is internal or external to the processor, for example, or composed of DRAM, SRAM, flash memory, or some other storage type, may depend on the available on-chip or off-chip storage space, the latency requirements of the training and / or inference functions being performed, the batch size of the data used in the inference and / or training of the neural network, or some combination of these factors.

[0035] In at least one embodiment, the inference and / or training logic 715 may include, but is not limited to, code and / or data storage 705 to store backpropagation and / or output weights and / or input / output data neural networks corresponding to neurons or layers of a neural network trained and / or used for inference in one or more embodiments. In at least one embodiment, during training and / or inference using one or more embodiments, the code and / or data storage 705 stores weight parameters and / or input / output data for each layer of a neural network trained or used in one or more embodiments during backpropagation of input / output data and / or weight parameters. In at least one embodiment, the training logic 715 may include or be coupled to code and / or data storage 705 for storing graph code or other software to control timing and / or sequence, wherein weight and / or other parameter information is loaded to configure logic including integer and / or floating-point units (collectively, an arithmetic logic unit (ALU)). In at least one embodiment, code (such as graph code) loads weight or other parameter information into the processor ALU based on the architecture of the neural network to which the code corresponds. In at least one embodiment, any portion of the code and / or data storage 705 may be included together with other on-chip or off-chip data storage, including the processor's L1, L2, or L3 cache or system memory. In at least one embodiment, any portion of the code and / or data storage 705 may be internal or external to one or more processors or other hardware logic devices or circuits. In at least one embodiment, the code and / or data storage 705 may be cache memory, DRAM, SRAM, non-volatile memory (e.g., flash memory), or other storage. In at least one embodiment, the choice between the code and / or data storage 705 being internal or external to the processor, for example, whether it consists of DRAM, SRAM, flash memory, or some other type of storage, depends on whether the available storage is on-chip or off-chip, the latency requirements of the training and / or inference functions being performed, the data batch size used in the inference and / or training of the neural network, or some combination of these factors.

[0036] In at least one embodiment, code and / or data storage 701 and code and / or data storage 705 may be separate storage structures. In at least one embodiment, code and / or data storage 701 and code and / or data storage 705 may be the same storage structure. In at least one embodiment, code and / or data storage 701 and code and / or data storage 705 may have partially identical storage structures and partially separate storage structures. In at least one embodiment, any portion of code and / or data storage 701 and code and / or data storage 705 may be included with other on-chip or off-chip data storage, including the processor's L1, L2, or L3 cache or system memory.

[0037] In at least one embodiment, the inference and / or training logic 715 may include, but is not limited to, one or more arithmetic logic units (“ALUs”) 710 (including integer and / or floating-point units) for performing logical and / or mathematical operations at least in part based on or instructed by training and / or inference code (e.g., graph code), the results of which may produce activations (e.g., output values ​​from layers or neurons within a neural network) stored in activation storage 720, which are functions of input / output and / or weight parameter data stored in code and / or data storage 701 and / or code and / or data storage 705. In at least one embodiment, activation is activated in response to execution instructions or other code, linear algebraic and / or matrix-based mathematical generation performed by ALU 710, and the activation is stored in activation storage 720, wherein weight values ​​stored in code and / or data storage 701 and / or code and / or data storage 705 are used as operands with other values, such as bias values, gradient information, momentum values, or other parameters or hyperparameters, and any or all of these can be stored in code and / or data storage 701 or code and / or data storage 705 or other on-chip or off-chip storage.

[0038] In at least one embodiment, one or more processors or other hardware logic devices or circuits include one or more ALUs 710, while in another embodiment, one or more ALUs 710 may be located outside the processor or other hardware logic device or the circuitry that uses them (e.g., a coprocessor). In at least one embodiment, one or more ALUs 710 may be included within an execution unit of a processor, or otherwise included in a group of ALUs accessible by the execution unit of the processor, which may be within the same processor or distributed among different processors of different types (e.g., a central processing unit, a graphics processing unit, a fixed-function unit, etc.). In at least one embodiment, code and / or data storage 701, code and / or data storage 705, and activation storage 720 may be the same processor or other hardware logic device or circuitry, while in another embodiment, they may be in different processors or other hardware logic devices or circuitry, or some combination of the same and different processors or other hardware logic devices or circuitry. In at least one embodiment, any portion of activation storage 720 may be included together with other on-chip or off-chip data storage, including the processor's L1, L2, or L3 cache or system memory. Furthermore, inference and / or training code may be stored together with other code accessible to the processor or other hardware logic or circuitry, and may be retrieved and / or processed using the processor’s fetch, decode, schedule, execute, exit, and / or other logic circuitry.

[0039] In at least one embodiment, the active memory 720 may be a cache memory, DRAM, SRAM, non-volatile memory (e.g., flash memory), or other memory. In at least one embodiment, the active memory 720 may be wholly or partially located inside or outside one or more processors or other logic circuits. In at least one embodiment, the choice of whether the active memory 720 is internal to or external to the processor may depend on the available on-chip or off-chip storage, the latency requirements for training and / or inference functions, the batch size of data used in inference and / or training the neural network, or some combination of these factors. For example, it may include DRAM, SRAM, flash memory, or other memory types. In at least one embodiment, Figure 7A The inference and / or training logic 715 shown can be used in conjunction with an application-specific integrated circuit (“ASIC”), such as a Tensorflow® processing unit from Google, an inference processing unit (IPU) from Graphcore™, or a Nervana® (e.g., “Lake Crest”) processor from Intel Corp. In at least one embodiment, Figure 7A The inference and / or training logic 715 shown can be used in conjunction with central processing unit (“CPU”) hardware, graphics processing unit (“GPU”) hardware, or other hardware such as field programmable gate array (“FPGA”).

[0040] Figure 7B Inference and / or training logic 715 according to at least one or more embodiments is illustrated. In at least one embodiment, the inference and / or training logic 715 may include, but is not limited to, hardware logic, wherein computational resources are dedicated or otherwise uniquely used in conjunction with weight values ​​or other information corresponding to one or more layers of neurons within a neural network. In at least one embodiment, Figure 7B The inference and / or training logic 715 shown can be used in conjunction with an application-specific integrated circuit (ASIC), such as a Tensorflow® processing unit from Google, an inference processing unit (IPU) from Graphcore™, or a Nervana® (e.g., “Lake Crest”) processor from Intel Corp. In at least one embodiment, Figure 7BThe inference and / or training logic 715 shown can be used in conjunction with central processing unit (CPU) hardware, graphics processing unit (GPU) hardware, or other hardware (e.g., field-programmable gate array (FPGA)). In at least one embodiment, the inference and / or training logic 715 includes, but is not limited to, code and / or data storage 701 and code and / or data storage 705, which can be used to store code (e.g., graph code), weight values, and / or other information, including bias values, gradient information, momentum values, and / or other parameter or hyperparameter information. Figure 7B In at least one embodiment shown, each of code and / or data storage 701 and code and / or data storage 705 is associated with dedicated computing resources (e.g., computing hardware 702 and computing hardware 706), respectively. In at least one embodiment, each of computing hardware 702 and computing hardware 706 includes one or more ALUs that perform mathematical functions (e.g., linear algebraic functions) only on the information stored in code and / or data storage 701 and code and / or data storage 705, respectively, and the results of the function execution are stored in activation storage 720.

[0041] In at least one embodiment, each of the code and / or data storage 701 and 705 and the corresponding computing hardware 702 and 706 corresponds to a different layer of the neural network, such that activation obtained from one “store / computation pair 701 / 702” of the code and / or data storage 701 and computing hardware 702 provides input as input to the next “store / computation pair 705 / 706” of the code and / or data storage 705 and computing hardware 706, in order to reflect the conceptual organization of the neural network. In at least one embodiment, each store / computation pair 701 / 702 and 705 / 706 may correspond to more than one neural network layer. In at least one embodiment, additional store / computation pairs (not shown) may be included in the inference and / or training logic 715 after or in parallel with the store / computation pairs 701 / 702 and 705 / 706.

[0042] Data Center Figure 8 An example data center 800 that can be used with at least one embodiment is shown. In at least one embodiment, the data center 800 includes a data center infrastructure layer 810, a framework layer 820, a software layer 830, and an application layer 840.

[0043] In at least one embodiment, such as Figure 8As shown, the data center infrastructure layer 810 may include a resource coordinator 812, packet computing resources 814, and node computing resources (“nodes CR”) 816(1)-816(N), where “N” represents any positive integer. In at least one embodiment, the nodes CR 816(1)-816(N) may include, but are not limited to, any number of central processing units (“CPUs”) or other processors (including accelerators, field-programmable gate arrays (FPGAs), graphics processors, etc.), memory devices (e.g., dynamic read-only memory), storage devices (e.g., solid-state drives or disk drives), network input / output (“NW I / O”) devices, network switches, virtual machines (“VMs”), power modules, and cooling modules, etc. In at least one embodiment, one or more of the nodes CR 816(1)-816(N) may be servers having one or more of the aforementioned computing resources.

[0044] In at least one embodiment, the grouped computing resource 814 may include individual groups (not shown) of node CRs housed within one or more racks, or a plurality of racks (also not shown) housed within data centers in various geographical locations. The individual groups of node CRs within the grouped computing resource 814 may include computing, networking, memory, or storage resources that can be configured or allocated to support groups of one or more workloads. In at least one embodiment, several node CRs, including CPUs or processors, may be grouped within one or more racks to provide computing resources to support one or more workloads. In at least one embodiment, the one or more racks may also include any number of power modules, cooling modules, and network switches, in any combination.

[0045] In at least one embodiment, resource coordinator 812 may be configured or otherwise control one or more nodes CR816(1)-816(N) and / or grouped computing resources 814. In at least one embodiment, resource coordinator 812 may include a Software Design Infrastructure (“SDI”) management entity for data center 800. In at least one embodiment, resource coordinator 812 may include hardware, software, or some combination thereof.

[0046] In at least one embodiment, such as Figure 8As shown, framework layer 820 includes a job scheduler 822, a configuration manager 824, a resource manager 826, and a distributed file system 828. In at least one embodiment, framework layer 820 may include a framework of software 832 supporting software layer 830 and / or one or more applications 842 supporting application layer 840. In at least one embodiment, software 832 or application 842 may respectively include web-based service software or applications, such as services or applications provided by Amazon Web Services, Google Cloud, and Microsoft Azure. In at least one embodiment, framework layer 820 may be, but is not limited to, a free and open-source software web application framework, such as Apache Spark™ (hereinafter referred to as "Spark") which can use distributed file system 828 for large-scale data processing (e.g., "big data"). In at least one embodiment, job scheduler 822 may include Spark drivers to facilitate the scheduling of workloads supported by the various layers of data center 800. In at least one embodiment, configuration manager 824 may be able to configure different layers, such as software layer 830 and framework layer 820 including Spark and distributed file system 828 for supporting large-scale data processing. In at least one embodiment, resource manager 826 is capable of managing cluster or group computing resources mapped to or allocated to support distributed file system 828 and job scheduler 822. In at least one embodiment, cluster or group computing resources may include group computing resources 814 on data center infrastructure layer 810. In at least one embodiment, resource manager 826 may coordinate with resource coordinator 812 to manage these mapped or allocated computing resources.

[0047] In at least one embodiment, the software 832 included in the software layer 830 may include software used by at least a portion of the nodes CR 816(1)-816(N), the grouped computing resources 814, and / or the distributed file system 828 of the framework layer 820. One or more types of software may include, but are not limited to, Internet web page search software, email virus scanning software, database software, and streaming video content software.

[0048] In at least one embodiment, one or more applications 842 included in application layer 840 may include one or more types of applications used by at least a portion of nodes CR 816(1)-816(N), grouped computing resources 814, and / or the distributed file system 828 of framework layer 820. One or more types of applications may include, but are not limited to, any number of genomics applications, cognitive computing and machine learning applications, including training or inference software, machine learning framework software (e.g., PyTorch, TensorFlow, Caffe, etc.), or other machine learning applications used in conjunction with one or more embodiments.

[0049] In at least one embodiment, any of the configuration manager 824, resource manager 826, and resource coordinator 812 can perform any number and type of self-modification actions based on any amount and type of data acquired in any technically feasible manner. In at least one embodiment, self-modification actions can mitigate potentially poor configuration decisions by data center operators of data center 800 and can prevent underutilization and / or poor performance of the data center.

[0050] In at least one embodiment, data center 800 may include tools, services, software, or other resources to train one or more machine learning models or to use one or more machine learning models to predict or infer information according to one or more embodiments described herein. For example, in at least one embodiment, a machine learning model can be trained by calculating weight parameters based on a neural network architecture using the software and computing resources described above with respect to data center 800. In at least one embodiment, information can be inferred or predicted using trained machine learning models corresponding to one or more neural networks using the resources described above with respect to data center 800 by using weight parameters calculated through one or more training techniques described herein.

[0051] In at least one embodiment, the data center may use a CPU, application-specific integrated circuit (ASIC), GPU, FPGA, or other hardware to utilize the aforementioned resources to perform training and / or inference. Furthermore, one or more of the aforementioned software and / or hardware resources may be configured as a service to allow a user to train or perform information inference, such as image recognition, speech recognition, or other artificial intelligence services.

[0052] Inference and / or training logic 715 is used to perform inference and / or training operations associated with one or more embodiments. The following is combined with... Figure 7A and / or Figure 7B Details are provided regarding the inference and / or training logic 715. In at least one embodiment, the inference and / or training logic 715 can be implemented in the system. Figure 8 Used in this context for inference or prediction operations based at least in part on weight parameters computed using neural network training operations, neural network functions and / or architectures, or neural network use cases described herein.

[0053] These components can be used to efficiently perform exponential or logarithmic approximations in half-precision floating-point format.

[0054] Computer System Figure 9 This is a block diagram illustrating an exemplary computer system 900 according to at least one embodiment. The exemplary computer system may be a system with interconnected devices and components, a system-on-a-chip (SoC), or some combination thereof formed with a processor, which may include an execution unit to execute instructions. In at least one embodiment, according to this disclosure, such as the embodiments described herein, the computer system 900 may include, but is not limited to, components such as processor 902, whose execution unit includes logic to execute algorithms for process data. In at least one embodiment, the computer system 900 may include a processor, such as the PENTIUM® processor family, Xeon™, Itanium®, XScale™ and / or StrongARM™, Intel® Core™ or Intel® Nervana™ microprocessors available from Intel Corporation of Santa Clara, California, although other systems (including PCs, engineering workstations, set-top boxes, etc.) may also be used. In at least one embodiment, the computer system 900 may execute a version of the Windows operating system available from Microsoft Corporation of Redmond, Wash., although other operating systems (such as UNIX and Linux), embedded software, and / or graphical user interfaces may also be used.

[0055] The embodiments can be used in other devices, such as handheld devices and embedded applications. Some examples of handheld devices include cellular phones, Internet Protocol (IP) devices, digital cameras, personal digital assistants (“PDAs”), and handheld PCs. In at least one embodiment, the embedded application may include a microcontroller, a digital signal processor (“DSP”), a system-on-a-chip (SoC), a network computer (“NetPC”), a set-top box, a network hub, a wide area network (“WAN”) switch, or any other system that can execute one or more instructions according to at least one embodiment.

[0056] In at least one embodiment, the computer system 900 may include, but is not limited to, a processor 902, which may include, but is not limited to, one or more execution units 908, to perform machine learning model training and / or inference according to the techniques described herein. In at least one embodiment, the computer system 900 is a single-processor desktop or server system, but in another embodiment, the computer system 900 may be a multiprocessor system. In at least one embodiment, the processor 902 may include, but is not limited to, a Complex Instruction Set Computing (“CISC”) microprocessor, a Reduced Instruction Set Computing (“RISC”) microprocessor, a Very Long Instruction Word (“VLIW”) microprocessor, a processor implementing instruction set combination, or any other processor device, such as a digital signal processor. In at least one embodiment, the processor 902 may be coupled to a processor bus 910, which may transmit data signals between the processor 902 and other components in the computer system 900.

[0057] In at least one embodiment, processor 902 may include, but is not limited to, a Level 1 (“L1”) internal cache memory (“cache”) 904. In at least one embodiment, processor 902 may have a single internal cache or multiple levels of internal cache. In at least one embodiment, cache memory 904 may reside outside of processor 902. Depending on specific implementation and requirements, other embodiments may also include a combination of internal and external caches. In at least one embodiment, register file 906 may store different types of data in various registers, including but not limited to integer registers, floating-point registers, status registers, and instruction pointer registers.

[0058] In at least one embodiment, an execution unit 908, including but not limited to logic for performing integer and floating-point operations, is also located within the processor 902. In at least one embodiment, the processor 902 may further include a microcode (“ucode”) read-only memory (“ROM”) for storing microcode for certain macro instructions. In at least one embodiment, the execution unit 908 may include logic for processing a packaged instruction set 909. In at least one embodiment, by including the packaged instruction set 909 in the instruction set of a general-purpose processor, along with the associated circuitry for executing the instructions, packaged data in the general-purpose processor 902 can be used to perform operations used by numerous multimedia applications. In one or more embodiments, many multimedia applications can be executed more quickly and efficiently by using the full width of the processor's data bus 910 to perform operations on the packaged data, which may eliminate the need to transfer smaller data units on the processor's data bus 910 to perform one or more operations on one data element at a time.

[0059] In at least one embodiment, execution unit 908 may also be used in a microcontroller, embedded processor, graphics device, DSP, and other types of logic circuitry. In at least one embodiment, computer system 900 may include, but is not limited to, memory 920. In at least one embodiment, memory 920 may be implemented as a dynamic random access memory (“DRAM”) device, a static random access memory (“SRAM”) device, a flash memory device, or other storage device. In at least one embodiment, memory 920 may store instructions 919 and / or data 921 represented by data signals that can be executed by processor 902.

[0060] In at least one embodiment, the system logic chip may be coupled to the processor bus 910 and the memory 920. In at least one embodiment, the system logic chip may include, but is not limited to, a memory controller hub (“MCH”) 916, and the processor 902 may communicate with the MCH 916 via the processor bus 910. In at least one embodiment, the MCH 916 may provide a high-bandwidth memory path 918 to the memory 920 for instruction and data storage, as well as for storage of graphics commands, data, and textures. In at least one embodiment, the MCH 916 may initiate data signals between the processor 902, the memory 920, and other components in the computer system 900, and bridge data signals between the processor bus 910, the memory 920, and the system I / O 922. In at least one embodiment, the system logic chip may provide a graphics port for coupling to a graphics controller. In at least one embodiment, the MCH 916 may be coupled to the memory 920 via the high-bandwidth memory path 918, and the graphics / video card 912 may be coupled to the MCH 916 via an Accelerated Graphics Port (“AGP”) interconnect 914.

[0061] In at least one embodiment, computer system 900 may use system I / O 922, a proprietary hub interface bus, for coupling MCH 916 to I / O controller hub (“ICH”) 930. In at least one embodiment, ICH 930 may provide direct connectivity to certain I / O devices via a local I / O bus. In at least one embodiment, the local I / O bus may include, but is not limited to, a high-speed I / O bus for connecting peripheral devices to memory 920, chipset, and processor 902. Examples may include, but are not limited to, an audio controller 929, a firmware hub (“Flash BIOS”) 928, a wireless transceiver 926, a data storage 924, a conventional I / O controller 923 including a user input and keyboard interface 925, a serial expansion port 927 (e.g., a Universal Serial Bus (USB) port), and a network controller 934. Data storage 924 may include a hard disk drive, floppy disk drive, CD-ROM device, flash memory device, or other mass storage device.

[0062] In at least one embodiment, Figure 9 A system including interconnected hardware devices or "chips" is shown, while in other embodiments, Figure 9 An exemplary system-on-a-chip (“SoC”) may be illustrated. In at least one embodiment, the device may be interconnected with a proprietary interconnect, a standardized interconnect (e.g., PCIe), or some combination thereof. In at least one embodiment, one or more components of the computer system 900 are interconnected using a compute fast link (CXL) interconnect.

[0063] The inference and / or training logic 715 is used to perform inference and / or training operations associated with one or more embodiments. This document combines... Figure 7A and / or Figure 7B Details are provided regarding the inference and / or training logic 715. In at least one embodiment, the inference and / or training logic 715 may be... Figure 9 Used in systems for reasoning or predicting operations based at least in part on weight parameters calculated using neural network training operations, neural network functions and / or architectures or neural network use cases described herein.

[0064] Such components can be used to efficiently perform exponential or logarithmic approximations in half-precision floating-point format.

[0065] Figure 10This is a block diagram illustrating an electronic device 1000 using a processor 1010 according to at least one embodiment. In at least one embodiment, the electronic device 1000 may be, for example, but not limited to, a laptop computer, tower server, rack server, blade server, laptop computer, desktop computer, tablet computer, mobile device, telephone, embedded computer, or any other suitable electronic device.

[0066] In at least one embodiment, the electronic device 1000 may, but is not limited to, a processor 1010 communicatively coupled to any suitable number or type of components, peripherals, modules, or devices. In at least one embodiment, the processor 1010 uses a bus or interface coupling, such as an I2C bus, a System Management Bus (“SMBus”), a Low Pin Count (LPC) bus, a Serial Peripheral Interface (“SPI”), a High Definition Audio (“HDA”) bus, a Serial Advanced Technology Attachment (“SATA”) bus, a Universal Serial Bus (“USB”) (versions 1, 2, and 3), or a Universal Asynchronous Receiver / Transmitter (“UART”) bus. In at least one embodiment, Figure 10 The system shown includes interconnected hardware devices or "chips," while in other embodiments, Figure 10 An exemplary system-on-chip (“SoC”) may be illustrated. In at least one embodiment, Figure 10 The device shown can be interconnected with proprietary interconnects, standardized interconnects (e.g., PCIe), or some combination thereof. In at least one embodiment, Figure 10 One or more components are interconnected using Computational Fast Link (CXL) interconnects.

[0067] In at least one embodiment, Figure 10 This may include a display 1024, a touchscreen 1025, a touchpad 1030, a near-field communication unit (“NFC”) 1045, a sensor hub 1040, a thermal sensor 1046, a fast chipset (“EC”) 1035, a trusted platform module (“TPM”) 1038, a BIOS / firmware / flash memory (“BIOS, FW Flash”) 1022, a DSP 1060, a drive 1020 (e.g., a solid-state drive (“SSD”) or a hard disk drive (“HDD”)), a wireless local area network unit (“WLAN”) 1050, a Bluetooth unit 1052, a wireless wide area network unit (“WWAN”) 1056, a global positioning system (GPS) unit 1055, a camera (“USB 3.0 camera”) 1054 (e.g., a USB 3.0 camera), and / or a low-power double data rate (“LPDDR”) memory unit (“LPDDR3”) 1015 implemented in, for example, the LPDDR3 standard. These components may each be implemented in any suitable manner.

[0068] In at least one embodiment, other components may be communicatively coupled to processor 1010 via the components discussed herein. In at least one embodiment, accelerometer 1041, ambient light sensor (“ALS”) 1042, compass 1043, and gyroscope 1044 may be communicatively coupled to sensor hub 1040. In at least one embodiment, thermal sensor 1039, fan 1037, keyboard 1036, and touchpad 1030 may be communicatively coupled to EC 1035. In at least one embodiment, speaker 1063, earphone 1064, and microphone (“mic”) 1065 may be communicatively coupled to audio unit (“audio codec and Class D amplifier”) 1062, which in turn may be communicatively coupled to DSP 1060. In at least one embodiment, audio unit 1062 may include, for example, but not limited to, audio encoder / decoder (“codec”) and Class D amplifier. In at least one embodiment, SIM card (“SIM”) 1057 may be communicatively coupled to WWAN unit 1056. In at least one embodiment, components such as WLAN unit 1050, Bluetooth unit 1052, and WWAN unit 1056 can be implemented as next-generation form factor (NGFF).

[0069] The inference and / or training logic 715 is used to perform inference and / or training operations associated with one or more embodiments. The following is in conjunction with... Figure 7A 7B and / or 7B provide details regarding the inference and / or training logic 715. In at least one embodiment, the inference and / or training logic 715 can be in the system Figure 10 It is used in the context of reasoning or predicting operations based at least in part on weight parameters calculated using neural network training operations, neural network functions and / or architectures or neural network use cases described herein.

[0070] Such components can be used to efficiently perform exponential or logarithmic approximations in half-precision floating-point format.

[0071] Figure 11 This is a block diagram of a processing system according to at least one embodiment. In at least one embodiment, system 1100 includes one or more processors 1102 and one or more graphics processors 1108, and may be a single-processor desktop system, a multi-processor workstation system, or a server system having a large number of processors 1102 or processor cores 1107. In at least one embodiment, system 1100 is a processing platform incorporated within a system-on-a-chip (SoC) integrated circuit for use in mobile, handheld, or embedded devices.

[0072] In at least one embodiment, system 1100 may include or be integrated into a server-based gaming platform, including a game console, mobile game console, handheld game console, or online game console, which are game and media consoles. In at least one embodiment, system 1100 is a mobile phone, smartphone, tablet computing device, or mobile internet device. In at least one embodiment, processing system 1100 may also include components coupled to or integrated into a wearable device, such as a smartwatch, smart glasses, augmented reality, or virtual reality device. In at least one embodiment, processing system 1100 is a television or set-top box device having one or more processors 1102 and a graphical interface generated by one or more graphics processors 1108.

[0073] In at least one embodiment, each of the one or more processors 1102 includes one or more processor cores 1107 for processing instructions that, when executed, perform operations against the system and user software. In at least one embodiment, each of the one or more processor cores 1107 is configured to process a particular instruction set 1109. In at least one embodiment, the instruction set 1109 may facilitate Complex Instruction Set Computing (CISC), Reduced Instruction Set Computing (RISC), or computation via Very Long Instruction Word (VLIW). In at least one embodiment, each processor core 1107 may process a different instruction set 1109, and the instruction sequence may include instructions that facilitate the emulation of other instruction sets. In at least one embodiment, the processor core 1107 may also include other processing devices, such as a digital signal processor (DSP).

[0074] In at least one embodiment, processor 1102 includes cache memory 1104. In at least one embodiment, processor 1102 may have a single internal cache or multiple levels of internal caches. In at least one embodiment, the cache memory is shared among various components of processor 1102. In at least one embodiment, processor 1102 also uses an external cache (e.g., a Level 3 (L3) cache or a last-level cache (LLC)) (not shown), which can be shared among processor cores 1107 using known cache alignment techniques. In at least one embodiment, processor 1102 further includes a register file 1106, which may include different types of registers (e.g., integer registers, floating-point registers, status registers, and instruction pointer registers) for storing different types of data. In at least one embodiment, register file 1106 may include general-purpose registers or other registers.

[0075] In at least one embodiment, one or more processors 1102 are coupled to one or more interface buses 1110 to transmit communication signals, such as address, data, or control signals, between the processors 1102 and other components in the system 1100. In at least one embodiment, the interface bus 1110 may be a processor bus, such as a version of the Direct Media Interface (DMI) bus. In at least one embodiment, the interface bus 1110 is not limited to the DMI bus and may include one or more peripheral component interconnect buses (e.g., PCI, PCI Express), memory buses, or other types of interface buses. In at least one embodiment, the processor 1102 includes an integrated memory controller 1116 and a platform controller hub 1130. In at least one embodiment, the memory controller 1116 facilitates communication between memory devices and other components of the system 1100, while the platform controller hub (PCH) 1130 provides connectivity to input / output (I / O) devices via a local I / O bus.

[0076] In at least one embodiment, memory device 1120 may be a dynamic random access memory (DRAM) device, a static random access memory (SRAM) device, a flash memory device, a phase-change memory device, or a device with suitable performance for use as processor memory. In at least one embodiment, memory device 1120 may be used as system memory of system 1100 to store data 1122 and instructions 1121 for use when one or more processors 1102 execute an application or process. In at least one embodiment, memory controller 1116 is also coupled to an optional external graphics processor 1112, which may communicate with one or more graphics processors 1108 of processor 1102 to perform graphics and media operations. In at least one embodiment, display device 1111 may be connected to processor 1102. In at least one embodiment, display device 1111 may include one or more internal display devices, such as in mobile electronic devices or laptop devices, or external display devices connected via a display interface (e.g., DisplayPort). In at least one embodiment, the display device 1111 may include a head-mounted display (HMD), such as a stereoscopic display device for virtual reality (VR) or augmented reality (AR) applications.

[0077] In at least one embodiment, the platform controller hub 1130 enables peripheral devices to connect to the storage device 1120 and the processor 1102 via a high-speed I / O bus. In at least one embodiment, the I / O peripheral devices include, but are not limited to, an audio controller 1146, a network controller 1134, a firmware interface 1128, a wireless transceiver 1126, a touch sensor 1125, and a data storage device 1124 (e.g., a hard disk drive, flash memory, etc.). In at least one embodiment, the data storage device 1124 may be connected via a storage interface (e.g., SATA) or via a peripheral bus, such as a peripheral component interconnect bus (e.g., PCI, PCIe). In at least one embodiment, the touch sensor 1125 may include a touchscreen sensor, a pressure sensor, or a fingerprint sensor. In at least one embodiment, the wireless transceiver 1126 may be a Wi-Fi transceiver, a Bluetooth transceiver, or a mobile network transceiver, such as a 3G, 4G, or LTE transceiver. In at least one embodiment, the firmware interface 1128 enables communication with the system firmware and may be, for example, a Unified Extensible Firmware Interface (UEFI). In at least one embodiment, network controller 1134 enables network connectivity to a wired network. In at least one embodiment, a high-performance network controller (not shown) is coupled to interface bus 1110. In at least one embodiment, audio controller 1146 is a multi-channel high-definition audio controller. In at least one embodiment, system 1100 includes an optional legacy I / O controller 1140 for coupling legacy (e.g., Personal System 2 (PS / 2)) devices to system 1100. In at least one embodiment, platform controller hub 1130 may also be connected to one or more Universal Serial Bus (USB) controllers 1142 that connect input devices, such as a keyboard and mouse combination 1143, a camera 1144, or other USB input devices.

[0078] In at least one embodiment, instances of the memory controller 1116 and platform controller hub 1130 may be integrated into a discrete external graphics processor, such as external graphics processor 1112. In at least one embodiment, the platform controller hub 1130 and / or the memory controller 1116 may be external to one or more processors 1102. For example, in at least one embodiment, system 1100 may include external memory controller 1116 and platform controller hub 1130, which may be configured as a memory controller hub and peripheral controller hub in a system chipset communicating with processor 1102.

[0079] Inference and / or training logic 715 is used to perform inference and / or training operations associated with one or more embodiments. This document combines... Figure 7A and / or Figure 7BDetails are provided regarding the inference and / or training logic 715. In at least one embodiment, some or all of the inference and / or training logic 715 may be incorporated into the graphics processor 1500. For example, in at least one embodiment, the training and / or inference techniques described herein may use one or more ALUs embodied in the graphics processor. Furthermore, in at least one embodiment, the inference and / or training operations described herein may use, in addition to Figure 7A and / or Figure 7B The logic is performed using logic other than that shown. In at least one embodiment, the weight parameters may be stored in on-chip or off-chip memory and / or registers (shown or not shown), which configure the ALU of the graphics processor to execute one or more machine learning algorithms, neural network architectures, use cases, or training techniques described herein.

[0080] Such components can be used to efficiently perform exponential or logarithmic approximations in half-precision floating-point format.

[0081] Figure 12 This is a block diagram of a processor 1200 according to at least one embodiment, the processor having one or more processor cores 1202A-1202N, an integrated memory controller 1214, and an integrated graphics processor 1208. In at least one embodiment, the processor 1200 may include additional cores, up to and including additional cores 1202N indicated by dashed boxes. In at least one embodiment, each processor core 1202A-1202N includes one or more internal cache units 1204A-1204N. In at least one embodiment, each processor core may also access one or more shared cache units 1206.

[0082] In at least one embodiment, internal cache units 1204A-1204N and shared cache unit 1206 represent a cache memory hierarchy within processor 1200. In at least one embodiment, cache memory units 1204A-1204N may include at least one level of instruction and data cache within each processor core and one or more levels of cache in a shared intermediate cache, such as Level 2 (L2), Level 3 (L3), Level 4 (L4), or other levels of cache, wherein the highest level of cache preceding external memory is classified as LLC. In at least one embodiment, cache alignment logic maintains alignment between the various cache units 1206 and 1204A-1204N.

[0083] In at least one embodiment, the processor 1200 may further include a set of one or more bus controller units 1216 and a system agent core 1210. In at least one embodiment, one or more bus controller units 1216 manage a set of peripheral buses, such as one or more PCI or PCIe buses. In at least one embodiment, the system agent core 1210 provides management functions for various processor components. In at least one embodiment, the system agent core 1210 includes one or more integrated memory controllers 1214 to manage access to various external memory devices (not shown).

[0084] In at least one embodiment, one or more processor cores 1202A-1202N include support for concurrent multithreading. In at least one embodiment, system agent core 1210 includes components for coordinating processor cores 1202A-1202N during multithreaded processing. In at least one embodiment, system agent core 1210 may additionally include a power control unit (PCU) including logic and components for regulating one or more power states of processor cores 1202A-1202N and graphics processor 1208.

[0085] In at least one embodiment, processor 1200 further includes a graphics processor 1208 for performing graph processing operations. In at least one embodiment, graphics processor 1208 is coupled to a shared cache unit 1206 and a system proxy core 1210 including one or more integrated memory controllers 1214. In at least one embodiment, system proxy core 1210 further includes a display controller 1211 for driving graphics processor outputs to one or more coupled displays. In at least one embodiment, display controller 1211 may also be a separate module coupled to graphics processor 1208 via at least one interconnect, or it may be integrated within graphics processor 1208.

[0086] In at least one embodiment, the ring-based interconnect unit 1212 is used to couple internal components of the processor 1200. In at least one embodiment, alternative interconnect units, such as point-to-point interconnects, switched interconnects, or other technologies, may be used. In at least one embodiment, the graphics processor 1208 is coupled to the ring-based interconnect unit 1212 via I / O link 1213.

[0087] In at least one embodiment, I / O link 1213 represents at least one of a variety of I / O interconnects, including packaged I / O interconnects that facilitate communication between various processor components and high-performance embedded memory module 1218 (e.g., eDRAM module). In at least one embodiment, each of processor cores 1202A-1202N and graphics processor 1208 uses embedded memory module 1218 as a shared last-level cache.

[0088] In at least one embodiment, processor cores 1202A-1202N are homogeneous cores executing a common instruction set architecture. In at least one embodiment, processor cores 1202A-1202N are heterogeneous in terms of instruction set architecture (ISA), with one or more processor cores 1202A-1202N executing a common instruction set, while one or more other processor cores 1202A-1202N execute a subset of the common instruction set or a different instruction set. In at least one embodiment, processor cores 1202A-1202N are heterogeneous in terms of microarchitecture, with one or more cores having relatively high power consumption coupled to one or more power cores having lower power consumption. In at least one embodiment, processor 1200 may be implemented on one or more chips or implemented as a SoC integrated circuit.

[0089] Inference and / or training logic 715 is used to perform inference and / or training operations associated with one or more embodiments. This document combines... Figure 7A and / or Figure 7B Details regarding the inference and / or training logic 715 are provided. In at least one embodiment, some or all of the inference and / or training logic 715 may be incorporated into the processor 1200. For example, in at least one embodiment, the training and / or inference techniques described herein may use one or more ALUs embodied in the graphics processor 1208, graphics cores 1202A-1202N, or... Figure 12 Among other components. Furthermore, in at least one embodiment, the inference and / or training operations described herein can use, except... Figure 7A and / or Figure 7B The logic is performed using logic other than that shown. In at least one embodiment, the weight parameters may be stored in on-chip or off-chip memory and / or registers (shown or not shown), which configure the ALU of the graphics processor 1200 to execute one or more machine learning algorithms, neural network architectures, use cases, or training techniques described herein.

[0090] Such components can be used to efficiently perform exponential or logarithmic approximations in half-precision floating-point format.

[0091] Virtualization computing platform Figure 13 This is an example data flow diagram of process 1300 for generating and deploying an image processing and inference pipeline according to at least one embodiment. In at least one embodiment, process 1300 can be deployed for use with imaging devices, processing devices, and / or other device types at one or more facilities 1302. Process 1300 can be executed within training system 1304 and / or deployment system 1306. In at least one embodiment, training system 1304 can be used to train, deploy, and implement machine learning models (e.g., neural networks, object detection algorithms, computer vision algorithms, etc.) for use in deployment system 1306. In at least one embodiment, deployment system 1306 can be configured to offload processing and computing resources between distributed computing environments to reduce infrastructure requirements at facility 1302. In at least one embodiment, one or more applications in the pipeline can use or invoke services of deployment system 1306 (e.g., inference, visualization, computation, AI, etc.) during application execution.

[0092] In at least one embodiment, some applications in the advanced processing and inference pipeline may use machine learning models or other AI to perform one or more processing steps. In at least one embodiment, the machine learning model may be trained at facility 1302 using data 1308 (such as imaging data) generated at facility 1302 (and stored on one or more Picture Archiving and Communication System (PACS) servers at facility 1302), or may be trained using imaging or sequencing data 1308 or a combination thereof from another facility (one or more facilities). In at least one embodiment, training system 1304 may be used to provide applications, services, and / or other resources for generating jobs, deployable machine learning models, for deployment system 1306.

[0093] In at least one embodiment, the model registry 1324 may be supported by an object storage system that supports versioning and object metadata. In at least one embodiment, the object storage may be accessed from within a cloud platform via, for example, a cloud storage-compatible application programming interface (API). In at least one embodiment, machine learning models within the model registry 1324 may be uploaded, listed, modified, or deleted by developers or partners of systems integrated with the API. In at least one embodiment, the API may provide access to methods that allow a user to associate a model with an application using appropriate credentials, enabling the model to be executed as part of the containerized instantiation of the application.

[0094] In at least one embodiment, training system 1304 ( Figure 13This could include a scenario where facility 1302 is training its own machine learning model, or has an existing machine learning model that needs to be optimized or updated. In at least one embodiment, imaging data 1308 generated by one or more imaging devices, sequencing devices, and / or other device types can be received. In at least one embodiment, once the imaging data 1308 is received, AI-assisted annotation 1310 can be used to help generate annotations corresponding to the imaging data 1308 for use as ground truth data for the machine learning model. In at least one embodiment, AI-assisted annotation 1310 can include one or more machine learning models (e.g., convolutional neural networks (CNNs)) that can be trained to generate annotations corresponding to certain types of imaging data 1308 (e.g., from certain devices). In at least one embodiment, AI-assisted annotation 1310 can then be used directly, or it can be adjusted or fine-tuned using annotation tools to generate ground truth data. In at least one embodiment, AI-assisted annotation 1310, labeled data 1312, or a combination thereof can be used as ground truth data for training the machine learning model. In at least one embodiment, the trained machine learning model may be referred to as output model 1316 and may be used by deployment system 1306 as described herein.

[0095] In at least one embodiment, the training pipeline may include scenarios where facility 1302 requires a machine learning model to perform one or more processing tasks of one or more applications in deployment system 1306, but facility 1302 may not currently have such a machine learning model (or may not have an optimized, efficient, or effective model for such a purpose). In at least one embodiment, an existing machine learning model may be selected from model registry 1324. In at least one embodiment, model registry 1324 may include machine learning models trained to perform various inference tasks on imaging data. In at least one embodiment, the machine learning models in model registry 1324 may have already been trained on imaging data from facilities other than facility 1302 (e.g., remote facilities). In at least one embodiment, the machine learning model may have already been trained on imaging data from one location, two locations, or any number of locations. In at least one embodiment, when training on imaging data from a particular location, training may occur at that location, or at least in a manner that protects the confidentiality of the imaging data or restricts the imaging data from being transmitted in the field. In at least one embodiment, once the model has been trained or partially trained at a location, the machine learning model may be added to model registry 1324. In at least one embodiment, the machine learning model can then be retrained or updated at any number of other facilities, and the retrained or updated model can be made available in model registry 1324. In at least one embodiment, the machine learning model can then be selected from model registry 1324—and referred to as output model 1316—and can be used in deployment system 1306 to perform one or more processing tasks for one or more applications of the deployment system.

[0096] In at least one embodiment, the scenario may include facility 1302, which requires a machine learning model to perform one or more processing tasks for one or more applications in deployment system 1306, but facility 1302 may not currently have such a machine learning model (or may not have an optimized, efficient, or effective model for such purposes). In at least one embodiment, the machine learning model selected from model registry 1324 may not be fine-tuned or optimized for the imaging data 1308 generated at facility 1302 due to population differences, robustness of training data used to train the machine learning model, anomalous diversity of training data, and / or other problems with the training data. In at least one embodiment, AI-assisted annotation 1310 may be used to help generate annotations corresponding to imaging data 1308, which is used as ground truth data for retraining or updating the machine learning model. In at least one embodiment, labeled data 1312 may be used as ground truth data for training the machine learning model. In at least one embodiment, retraining or updating the machine learning model may be referred to as model training 1314. In at least one embodiment, model training 1314 (e.g., AI-assisted annotation 1310, labeled data 1312, or a combination thereof) can be used as ground truth data for retraining or updating the machine learning model. In at least one embodiment, the trained machine learning model may be referred to as output model 1316 and may be used by deployment system 1306 as described herein.

[0097] In at least one embodiment, deployment system 1306 may include software 1318, service 1320, hardware 1322, and / or other components, features, and functions. In at least one embodiment, deployment system 1306 may include a software "stack" such that software 1318 can be built on top of service 1320 and can be used to perform some or all of the processing tasks, and service 1320 and software 1318 can be built on top of hardware 1322 and use hardware 1322 to perform the processing, storage, and / or other computational tasks of deployment system 1306. In at least one embodiment, software 1318 may include any number of different containers, each of which can perform an instantiation of an application. In at least one embodiment, each application can perform one or more processing tasks (e.g., inference, object detection, feature detection, segmentation, image enhancement, calibration, etc.) in a high-level processing and inference pipeline. In at least one embodiment, in addition to receiving and configuring imaging data for use by each container and / or by facility 1302 after processing through the pipeline, advanced processing and inference pipelines (e.g., to convert output back to available data types) can be defined based on the selection of different containers desired or required for processing imaging data 1308. In at least one embodiment, a combination of containers within software 1318 (e.g., constituting a pipeline) may be referred to as a virtual instrument (as described in more detail herein), and the virtual instrument may utilize service 1320 and hardware 1322 to perform some or all of the processing tasks of an application instantiated within the container.

[0098] In at least one embodiment, the data processing pipeline may receive input data (e.g., imaging data 1308) in a specific format in response to an inference request (e.g., a request from a user of deployment system 1306). In at least one embodiment, the input data may represent one or more images, videos, and / or other data representations generated by one or more imaging devices. In at least one embodiment, the data may be preprocessed as part of the data processing pipeline to prepare it for processing by one or more applications. In at least one embodiment, post-processing may be performed on the output of one or more inference tasks or other processing tasks of the pipeline to prepare output data for the next application and / or to prepare output data for user transmission and / or use (e.g., as a response to an inference request). In at least one embodiment, the inference task may be performed by one or more machine learning models, such as trained or deployed neural networks, which may include the output model 1316 of training system 1304.

[0099] In at least one embodiment, the tasks of the data processing pipeline can be encapsulated in containers, each container representing a discrete, fully functional instantiation of an application and a virtualized computing environment capable of referencing a machine learning model. In at least one embodiment, containers or applications can be published to a private (e.g., limited access) area of ​​a container registry (described in more detail herein), and trained or deployed models can be stored in a model registry 1324 and associated with one or more applications. In at least one embodiment, an image of the application (e.g., a container image) can be used in the container registry, and once a user selects an image from the container registry for deployment in the pipeline, that image can be used to generate containers for instantiation of the application for use by the user's system.

[0100] In at least one embodiment, a developer (e.g., a software developer, clinician, physician, etc.) can develop, publish, and store an application (e.g., as a container) for performing image processing and / or inference on provided data. In at least one embodiment, a software development kit (SDK) associated with the system can be used to perform development, publication, and / or storage (e.g., to ensure the developed application and / or container conforms to or is compatible with the system). In at least one embodiment, the developed application can be tested locally using the SDK (e.g., at a first facility, testing data from a first facility), the SDK serving as a system (e.g.,...). Figure 12 System 1200 may support at least some services 1320. In at least one embodiment, since a DICOM object may contain one to hundreds of images or other data types, and due to variations in the data, the developer may be responsible for managing (e.g., setting up constructs for preprocessing built into the application, etc.) the extraction and preparation of incoming data. In at least one embodiment, once verified by system 1300 (e.g., for accuracy), the application becomes available in the container registry for user selection and / or implementation to perform one or more processing tasks on data at the user's facility (e.g., a second facility).

[0101] In at least one embodiment, the developer can then share the application or container over a network for the system (e.g., Figure 13The system 1300 allows for user access and use. In at least one embodiment, completed and validated applications or containers may be stored in a container registry, and associated machine learning models may be stored in a model registry 1324. In at least one embodiment, a requesting entity (which provides an inference or image processing request) may browse the container registry and / or model registry 1324 to obtain applications, containers, datasets, machine learning models, etc., select desired combinations of elements to include in the data processing pipeline, and submit an image processing request. In at least one embodiment, the request may include input data necessary to execute the request (and, in some examples, patient-related data), and / or may include selections of applications and / or machine learning models to be executed when the request is processed. In at least one embodiment, the request may then be passed to one or more components of the deployment system 1306 (e.g., the cloud) to perform processing in the data processing pipeline. In at least one embodiment, processing performed by the deployment system 1306 may include referencing elements (e.g., applications, containers, models, etc.) selected from the container registry and / or model registry 1324. In at least one embodiment, once the results are generated through the pipeline, the results can be returned to the user for reference (e.g., for viewing in a suite of viewing applications executed locally, on a local workstation, or on a terminal).

[0102] In at least one embodiment, service 1320 may be utilized to assist in processing or executing applications or containers in the pipeline. In at least one embodiment, service 1320 may include computing services, artificial intelligence (AI) services, visualization services, and / or other service types. In at least one embodiment, service 1320 may provide functionality common to one or more applications in software 1318, thus abstracting functionality into services that can be invoked or utilized by applications. In at least one embodiment, the functionality provided by service 1320 can operate dynamically and more efficiently, while also allowing applications to process data in parallel (e.g., using parallel computing platform 1230). Figure 12This allows for good scaling. In at least one embodiment, it is not required that every application providing the same functionality as service 1320 must have a corresponding instance of service 1320, but rather service 1320 can be shared between and among various applications. In at least one embodiment, as a non-limiting example, the service may include an inference server or engine that can be used to perform detection or segmentation tasks. In at least one embodiment, a model training service may be included, which can provide the ability to train and / or retrain machine learning models. In at least one embodiment, a data augmentation service may be further included, which can provide GPU-accelerated data (e.g., DICOM, RIS, CIS, conforming to REST, RPC, raw, etc.) extraction, resizing, scaling, and / or other enhancements. In at least one embodiment, a visualization service may be used, which can add image rendering effects (e.g., ray tracing, rasterization, denoising, sharpening, etc.) to add realism to two-dimensional (2D) and / or three-dimensional (3D) models. In at least one embodiment, a virtual instrument service may be included, which provides beamforming, segmentation, inference, imaging, and / or support for other applications within the virtual instrument pipeline.

[0103] In at least one embodiment, where service 1320 includes an AI service (e.g., an inference service), as part of application execution, one or more machine learning models can be executed by invoking (e.g., as an API call) the inference service (e.g., an inference server) to execute one or more machine learning models or their processing. In at least one embodiment, where another application includes one or more machine learning models for a segmentation task, the application can invoke the inference service to execute the machine learning models for performing one or more processing operations associated with the segmentation task. In at least one embodiment, software 1318 implementing advanced processing and inference pipelines, including a segmentation application and an anomaly detection application, can be pipelined because each application can invoke the same inference service to execute one or more inference tasks.

[0104] In at least one embodiment, hardware 1322 may include a GPU, CPU, graphics card, AI / deep learning system (e.g., an AI supercomputer, such as NVIDIA's DGX), cloud platform, or a combination thereof. In at least one embodiment, different types of hardware 1322 may be used to provide efficient, specially built support for software 1318 and services 1320 in deployment system 1306. In at least one embodiment, GPU processing may be used to perform local processing (e.g., at facility 1302) within the AI / deep learning system, in the cloud system, and / or other processing components of deployment system 1306 to improve the efficiency, accuracy, and performance of image processing and generation. In at least one embodiment, as a non-limiting example, software 1318 and / or services 1320 may be optimized for GPU processing in relation to deep learning, machine learning, and / or high-performance computing. In at least one embodiment, at least some of the computing environment of deployment system 1306 and / or training system 1304 may be executed in a data center, one or more supercomputers, or high-performance computing systems with GPU-optimized software (e.g., a hardware and software combination of an NVIDIA DGX system). In at least one embodiment, as described herein, hardware 1322 may include any number of GPUs that can be invoked to perform data processing in parallel. In at least one embodiment, the cloud platform may also include GPU-optimized execution for deep learning tasks, GPU processing for machine learning tasks, or other computational tasks. In at least one embodiment, an AI / deep learning supercomputer and / or GPU-optimized software (e.g., as provided on NVIDIA's DGX systems) may be used as a hardware abstraction and scaling platform to execute the cloud platform (e.g., NVIDIA's NGC). In at least one embodiment, the cloud platform may integrate application container cluster systems or coordination systems (e.g., Kubernetes) across multiple GPUs to enable seamless scaling and load balancing.

[0105] Figure 14 This is a system diagram of an example system 1400 for generating and deploying an imaging deployment pipeline according to at least one embodiment. In at least one embodiment, system 1400 can be used to implement Figure 13 The process 1300 and / or other processes include advanced processing and inference pipelines. In at least one embodiment, system 1400 may include training system 1304 and deployment system 1306. In at least one embodiment, training system 1304 and deployment system 1306 may be implemented using software 1318, service 1320 and / or hardware 1322, as described herein.

[0106] In at least one embodiment, system 1400 (e.g., training system 1304 and / or deployment system 1306) may be implemented in a cloud computing environment (e.g., using cloud 1426). In at least one embodiment, system 1400 may be implemented locally (in relation to a healthcare facility) or as a combination of cloud computing resources and local computing resources. In at least one embodiment, access to the API in cloud 1426 may be restricted to authorized users by establishing security measures or protocols. In at least one embodiment, the security protocol may include network tags, which may be signed by an authentication service (e.g., AuthN, AuthZ, Gluecon, etc.) and may carry appropriate authorization. In at least one embodiment, the API of the virtual instrument (described herein) or other instances of system 1400 may be restricted to a set of public IPs that have been audited or authorized for interaction.

[0107] In at least one embodiment, the various components of system 1400 may communicate with each other using any of a variety of different network types, including but not limited to local area networks (LANs) and / or wide area networks (WANs) via wired and / or wireless communication protocols. In at least one embodiment, communication between facilities and components of system 1400 (e.g., for sending inference requests, for receiving the results of inference requests, etc.) may be transmitted via one or more data buses, wireless data protocols (Wi-Fi), wired data protocols (e.g., Ethernet), etc.

[0108] In at least one embodiment, similar to the description herein. Figure 13 As described, training system 1304 can execute training pipeline 1404. In at least one embodiment, where deployment system 1306 uses one or more machine learning models in deployment pipeline 1410, training pipeline 1404 can be used to train or retrain one or more (e.g., pre-trained) models, and / or implement one or more pre-trained models 1406 (e.g., without retraining or updating). In at least one embodiment, as a result of training pipeline 1404, output model 1316 can be generated. In at least one embodiment, training pipeline 1404 can include any number of processing steps, such as, but not limited to, transformation or adaptation of imaging data (or other input data). In at least one embodiment, different training pipelines 1404 can be used for different machine learning models used by deployment system 1306. In at least one embodiment, similar to the description of... Figure 13 The training pipeline 1404 described in the first example can be used for the first machine learning model, similar to the one described above. Figure 13 The training pipeline 1404 described in the second example can be used for a second machine learning model, similar to the one described above. Figure 13The training pipeline 1404 of the third example described can be used for a third machine learning model. In at least one embodiment, any combination of tasks within the training system 1304 can be used according to the requirements of each respective machine learning model. In at least one embodiment, one or more machine learning models may have already been trained and are ready for deployment, so the training system 1304 may not perform any processing on the machine learning models, and one or more machine learning models may be implemented by the deployment system 1306.

[0109] In at least one embodiment, depending on the implementation or embodiment, the output model 1316 and / or the pre-trained model 1406 may include any type of machine learning model. In at least one embodiment, and not limited thereto, the machine learning model used by system 1400 may include models using linear regression, logistic regression, decision trees, support vector machines (SVM), Naive Bayes, k-nearest neighbors (Knn), k-means clustering, random forests, dimensionality reduction algorithms, gradient boosting algorithms, neural networks (e.g., autoencoders, convolutions, recursion, perceptrons, long / short-term memory (LSTM), Hopfield, Boltzmann, deep belief, deconvolution, generative adversarial, liquid state machines, etc.), and / or other types of machine learning models.

[0110] In at least one embodiment, the training pipeline 1404 may include AI-assisted annotations, as described herein regarding at least Figure 14More specifically, in at least one embodiment, the labeled data 1312 can be generated using any number of techniques (e.g., conventional annotations). In at least one embodiment, in some examples, labels or other annotations can be generated by drawing programs (e.g., annotation programs), computer-aided design (CAD) programs, tagging programs, another type of application suitable for generating truth values ​​or labels, and / or can be hand-drawn. In at least one embodiment, truth data can be synthetically generated (e.g., generated from computer models or renderings), realistically generated (e.g., designed and generated from real-world data), automatically generated by machines (e.g., extracting features from data using feature analysis and learning, and then generating labels), manually annotated (e.g., taggers or annotation experts, defining the placement of labels), and / or combinations thereof. In at least one embodiment, for each instance of imaging data 1308 (or other data types used by machine learning models), there can be corresponding truth data generated by training system 1304. In at least one embodiment, AI-assisted annotation can be performed as part of deployment pipeline 1410; supplementing or replacing AI-assisted annotations included in training pipeline 1404. In at least one embodiment, system 1400 may include a multi-layer platform, which may include a software layer (e.g., software 1318) of a diagnostic application (or other application type) capable of performing one or more medical imaging and diagnostic functions. In at least one embodiment, system 1400 may be communicatively coupled (e.g., via an encrypted link) to a network of PACS servers in one or more facilities. In at least one embodiment, system 1400 may be configured to access and reference data from PACS servers to perform operations such as training machine learning models, deploying machine learning models, image processing, inference, and / or other operations.

[0111] In at least one embodiment, the software layer may be implemented as a secure, encrypted, and / or certified API through which an application or container can be invoked (e.g., called) from an external environment (e.g., facility 1302). In at least one embodiment, the application may then invoke or execute one or more services 1320 to perform computational, AI, or visualization tasks associated with their respective applications, and the software 1318 and / or service 1320 may utilize hardware 1322 to perform processing tasks efficiently and effectively. In at least one embodiment, a pair of DICOM adapters 1402A, 1402B may be used for communication sent to or received by the training system 1304 and deployment system 1306.

[0112] In at least one embodiment, deployment system 1306 may execute deployment pipeline 1410. In at least one embodiment, deployment pipeline 1410 may include any number of applications, which may be sequential, non-sequential, or otherwise applied to imaging data (and / or other data types) – including AI-assisted annotation, the imaging data being generated by imaging devices, sequencing devices, genomics devices, etc., as described above. In at least one embodiment, as described herein, deployment pipeline 1410 for an individual device may be referred to as a virtual instrument for the device (e.g., a virtual ultrasound instrument, a virtual CT scanner, a virtual sequencing instrument, etc.). In at least one embodiment, for a single device, more than one deployment pipeline 1410 may exist, depending on the desired information from the data generated from the device. In at least one embodiment, a first deployment pipeline 1410 may exist if it is desired to detect an anomaly from an MRI machine, and a second deployment pipeline 1410 may exist if it is desired to perform image enhancement from the output of the MRI machine.

[0113] In at least one embodiment, the image generation application may include processing tasks that include using a machine learning model. In at least one embodiment, a user may wish to use their own machine learning model or select a machine learning model from the model registry 1324. In at least one embodiment, a user may implement their own machine learning model or select a machine learning model to be included in the application performing the processing task. In at least one embodiment, the application may be optional and customizable, and by defining the construction of the application, the deployment and implementation of the application for a particular user is presented as a more seamless user experience. In at least one embodiment, by leveraging other features of system 1400 (e.g., service 1320 and hardware 1322), deployment pipeline 1410 can be more user-friendly, provide easier integration, and produce more accurate, efficient, and timely results.

[0114] In at least one embodiment, deployment system 1306 may include a user interface (“UI”) 1414 (e.g., a graphical user interface, a web interface, etc.) that can be used to select applications to be included in deployment pipeline 1410, deploy applications, modify or change applications or their parameters or configurations, use and interact with deployment pipeline 1410 during setup and / or deployment, and / or otherwise interact with deployment system 1306. In at least one embodiment, although not shown with respect to training system 1304, UI 1414 (or different user interfaces) can be used to select models to be used in deployment system 1306, to select models to be trained or retrained in training system 1304, and / or to otherwise interact with training system 1304.

[0115] In at least one embodiment, in addition to the application coordination system 1428, a pipeline manager 1412 may be used to manage interactions between applications or containers deployed through the pipeline 1410 and services 1320 and / or hardware 1322. In at least one embodiment, the pipeline manager 1412 may be configured to facilitate interactions from application to application, from application to service 1320, and / or from application or service to hardware 1322. In at least one embodiment, although shown as included in software 1318, this is not intended to be limiting, and in some examples, the pipeline manager 1412 may be included in service 1320. In at least one embodiment, the application coordination system 1428 (e.g., Kubernetes, DOCKER, etc.) may include a container coordination system that can group applications into containers as logical units for coordination, management, scaling, and deployment. In at least one embodiment, by associating applications from the deployment pipeline 1410 (e.g., rebuilding applications, splitting applications, etc.) with individual containers, each application can execute in a self-contained environment (e.g., at the kernel level) to improve speed and efficiency.

[0116] In at least one embodiment, each application and / or container (or its image) can be developed, modified, and deployed independently (e.g., a first user or developer can develop, modify, and deploy a first application, and a second user or developer can develop, modify, and deploy a second application separate from the first user or developer). This allows focus on the tasks of a single application and / or container without being hindered by the tasks of another application or container. In at least one embodiment, the pipeline manager 1412 and the application coordination system 1428 can facilitate communication and collaboration between different containers or applications. In at least one embodiment, the application coordination system 1428 and / or the pipeline manager 1412 can facilitate communication and resource sharing between and within each application or container, provided that the expected inputs and / or outputs of each container or application are known to the system (e.g., based on the application or container's architecture). In at least one embodiment, since one or more applications or containers in the deployment pipeline 1410 can share the same services and resources, the application coordination system 1428 can coordinate, load balance, and determine the sharing of services or resources between and within the various applications or containers. In at least one embodiment, the scheduler can be used to track the resource requirements of applications or containers, the current or planned use of these resources, and resource availability. Therefore, in at least one embodiment, the scheduler can allocate resources to different applications and distribute resources between and among applications, taking into account the system's needs and availability. In some examples, the scheduler (and / or other components of the application coordination system 1428) can determine resource availability and distribution based on constraints imposed on the system (e.g., user constraints), such as Quality of Service (QoS), the urgency of data output (e.g., to determine whether to perform real-time processing or delayed processing), etc.

[0117] In at least one embodiment, service 1320, utilized and shared by applications or containers in deployment system 1306, may include computing service 1416, AI service 1418, visualization service 1420, and / or other service types. In at least one embodiment, an application may invoke (e.g., execute) one or more services 1320 to perform processing operations for the application. In at least one embodiment, an application may utilize computing service 1416 to perform supercomputing or other high-performance computing (HPC) tasks. In at least one embodiment, one or more computing services 1416 may be utilized to perform parallel processing (e.g., using parallel computing platform 1430) to process data substantially simultaneously through one or more applications and / or one or more tasks of a single application. In at least one embodiment, parallel computing platform 1430 (e.g., NVIDIA's CUDA) enables general-purpose computing on GPUs (GPGPUs) (e.g., GPU / Graphics 1422). In at least one embodiment, the software layer of parallel computing platform 1430 may provide access to the GPU's virtual instruction set and parallel computing elements to execute computational kernels. In at least one embodiment, the parallel computing platform 1430 may include memory, and in some embodiments, memory may be shared between and within multiple containers, and / or between and within different processing tasks within a single container. In at least one embodiment, inter-process communication (IPC) calls may be generated for multiple containers and / or multiple processes within containers to use the same data from a shared memory segment of the parallel computing platform 1430 (e.g., where multiple different stages of one or more applications are processing the same information). In at least one embodiment, instead of copying data and moving it to different locations in memory (e.g., read / write operations), the same data in the same memory location can be used for any number of processing tasks (e.g., at the same time, at different times, etc.). In at least one embodiment, this information about the new location of the data can be stored and shared between applications because the resulting data from processing is used to generate new data. In at least one embodiment, the location of the data, and the location of the updated or modified data, may be part of the definition of how the payload in the container is understood.

[0118] In at least one embodiment, AI service 1418 may be used to perform an inference service for executing a machine learning model associated with the application (e.g., a task to perform one or more processing tasks of the application). In at least one embodiment, AI service 1418 may utilize AI system 1424 to execute a machine learning model (e.g., a neural network such as a CNN) for segmentation, reconstruction, object detection, feature detection, classification, and / or other inference tasks. In at least one embodiment, the application deploying pipeline 1410 may use one or more output models 1316 from training system 1304 and / or other models of the application to perform inference on imaging data. In at least one embodiment, two or more examples of using application coordination system 1428 (e.g., a scheduler) for inference may be available. In at least one embodiment, a first category may include a high-priority / low-latency path that can implement a higher service level protocol, such as for performing inference on urgent requests in emergency situations or for radiologists during diagnostic procedures. In at least one embodiment, a second category may include a standard priority path that can be used for requests that may not be urgent or for situations where analysis can be performed at a later time. In at least one embodiment, the application coordination system 1428 may allocate resources (e.g., services 1320 and / or hardware 1322) based on priority paths for different inference tasks of the AI ​​service 1418.

[0119] In at least one embodiment, shared memory may be installed into AI service 1418 in system 1400. In at least one embodiment, shared memory may operate as a cache (or other storage device type) and may be used to process inference requests from applications. In at least one embodiment, when an inference request is submitted, a set of API instances of deployment system 1306 may receive the request and may select one or more instances (e.g., for best fit, for load balancing, etc.) to process the request. In at least one embodiment, to process the request, the request may be fed into a database, and if not already in the cache, a machine learning model may be located from model registry 1324. A verification step may ensure that an appropriate machine learning model is loaded into the cache (e.g., shared memory), and / or a copy of the model may be saved to the cache. In at least one embodiment, if the application is not already running or there are not enough instances of the application, a scheduler (e.g., the scheduler of pipeline manager 1412) may be used to start the application referenced in the request. In at least one embodiment, if an inference server has not yet been started to execute the model, an inference server may be started. Any number of inference servers may be started for each model. In at least one embodiment, in a pull model that clusters inference servers, the model can be cached whenever load balancing is favorable. In at least one embodiment, the inference servers can be statically loaded into the corresponding distributed servers.

[0120] In at least one embodiment, an inference server running in a container can be used to perform inference. In at least one embodiment, an instance of the inference server can be associated with a model (and optionally multiple versions of the model). In at least one embodiment, if an instance of the inference server does not exist when a request to perform inference on the model is received, a new instance can be loaded. In at least one embodiment, when the inference server is started, a model can be passed to the inference server, allowing the same container to be used to serve different models, as long as the inference server runs as different instances.

[0121] In at least one embodiment, during application execution, an inference request for a given application can be received, and a container (e.g., an instance of a hosted inference server) can be loaded (if not already loaded), and a launcher can be invoked. In at least one embodiment, preprocessing logic within the container can (e.g., using a CPU and / or GPU) load, decode, and / or perform any additional preprocessing on the incoming data. In at least one embodiment, once the data is ready for inference, the container can infer the data as needed. In at least one embodiment, this can include a single inference call for an image (e.g., a hand X-ray) or can request inference for hundreds of images (e.g., a chest CT scan). In at least one embodiment, the application can summarize the results before completion, which may include, but is not limited to, a single confidence score, pixel-level segmentation, voxel-level segmentation, generating visualizations, or generating text to summarize the results. In at least one embodiment, different priorities can be assigned to different models or applications. For example, some models may have a real-time (TAT less than 1 minute) priority, while other models may have a lower priority (e.g., TAT less than 10 minutes). In at least one embodiment, model execution time can be measured from the requesting agency or entity, and may include cooperative network traversal time and inference service execution time.

[0122] In at least one embodiment, the transfer of requests between service 1320 and the inference application can be hidden behind a software development kit (SDK) and robust transfer can be provided via queues. In at least one embodiment, requests are placed in queues via an API for individual application / tenant ID combinations, and the SDK pulls requests from the queues and provides them to the application. In at least one embodiment, the name of the queue can be provided in the environment where the SDK picks up the queue. In at least one embodiment, asynchronous communication via queues may be useful because it allows any instance of the application to pick up work when it becomes available. Results can be sent back via queues to ensure no data loss. In at least one embodiment, queues can also provide the ability to partition work, as the highest priority work can go into a queue connected to a majority of instances of the application, while the lowest priority work can go into a queue connected to a single instance that processes tasks in the order they are received. In at least one embodiment, the application can run on a GPU-accelerated instance generated in cloud 1426, and the inference service can perform inference on the GPU.

[0123] In at least one embodiment, visualization service 1420 can be used to generate visualizations for viewing the output of application and / or deployment pipeline 1410. In at least one embodiment, visualization service 1420 can utilize GPU / Graphic 1422 to generate visualizations. In at least one embodiment, visualization service 1420 can implement rendering effects such as ray tracing to generate higher quality visualizations. In at least one embodiment, visualizations can include, but are not limited to, 2D image rendering, 3D volume rendering, 3D volume reconstruction, 2D tomographic slicing, virtual reality display, augmented reality display, etc. In at least one embodiment, a virtualized environment can be used to generate virtual interactive displays or environments (e.g., virtual environments) for system users (e.g., doctors, nurses, radiologists, etc.) to interact with. In at least one embodiment, visualization service 1420 can include an internal visualizer, cinematic and / or other rendering or image processing capabilities or functions (e.g., ray tracing, rasterization, internal optics, etc.).

[0124] In at least one embodiment, hardware 1322 may include GPU / Graphic 1422, AI system 1424, cloud 1426, and / or any other hardware for performing training system 1304 and / or deployment system 1306. In at least one embodiment, GPU / Graphic 1422 (e.g., NVIDIA's TESLA and / or QUADRO GPUs) may include any number of GPUs that can be used to perform processing tasks for any feature or function of computing service 1416, AI service 1418, visualization service 1420, other services, and / or software 1318. For example, for AI service 1418, GPU / Graphic 1422 may be used to perform preprocessing on imaging data (or other data types used by machine learning models), postprocessing on the output of machine learning models, and / or inference (e.g., to execute machine learning models). In at least one embodiment, cloud 1426, AI system 1424, and / or other components of system 1400 may use GPU / Graphic 1422. In at least one embodiment, cloud 1426 may include a GPU-optimized platform for deep learning tasks. In at least one embodiment, AI system 1424 may use a GPU, and one or more AI systems 1424 may be used to perform cloud 1426 (or at least part of a task for deep learning or inference). Similarly, although hardware 1322 is shown as a discrete component, this is not intended to be limiting, and any component of hardware 1322 may be combined with or utilized by any other component of hardware 1322.

[0125] In at least one embodiment, AI system 1424 may include a specially built computing system (e.g., a supercomputer or HPC) configured for inference, deep learning, machine learning, and / or other artificial intelligence tasks. In at least one embodiment, in addition to CPU, RAM, memory, and / or other components, features, or functions, AI system 1424 (e.g., NVIDIA's DGX) may also include GPU-optimized software (e.g., a software stack) that can be performed using multiple GPUs / Graphics 1422. In at least one embodiment, one or more AI systems 1424 may be implemented in a cloud 1426 (e.g., in a data center) to perform some or all of the AI-based processing tasks of system 1400.

[0126] In at least one embodiment, cloud 1426 may include GPU-accelerated infrastructure (e.g., NVIDIA's NGC) that can provide a GPU-optimized platform for performing processing tasks of system 1400. In at least one embodiment, cloud 1426 may include AI system 1424 for performing one or more AI-based tasks of system 1400 (e.g., as a hardware abstraction and scaling platform). In at least one embodiment, cloud 1426 may be integrated with application coordination system 1428 that utilizes multiple GPUs to enable seamless scaling and load balancing between and within applications and services 1320. In at least one embodiment, as described herein, cloud 1426 may be responsible for performing at least some of the services 1320 of system 1400, including computing service 1416, AI service 1418, and / or visualization service 1420. In at least one embodiment, cloud 1426 may perform large and small batch inference (e.g., perform NVIDIA's TENSOR RT), provide accelerated parallel computing APIs and platform 1430 (e.g., NVIDIA's CUDA), perform application coordination system 1428 (e.g., KUBERNETES), provide graphics rendering APIs and platform (e.g., for ray tracing, 2D graphics, 3D graphics and / or other rendering techniques to produce higher quality cinematic effects), and / or may provide other functionalities for system 1400.

[0127] Figure 15A A data flow diagram of a process 1500 for training, retraining, or updating a machine learning model according to at least one embodiment is shown. In at least one embodiment, a non-limiting example can be used. Figure 14 The system 1400 executes the process 1500. In at least one embodiment, the process 1500 may utilize services and / or hardware as described herein. In at least one embodiment, the refined model 1512 generated by the process 1500 may be executed by a deployment system for one or more containerized applications in the deployment pipeline.

[0128] In at least one embodiment, model training 1514 may include retraining or updating the initial model 1504 (e.g., a pre-trained model) using new training data (e.g., new input data, such as customer dataset 1506, and / or new ground truth data associated with the input data). In at least one embodiment, to retrain or update the initial model 1504, the output or loss layer of the initial model 1504 may be reset, deleted, and / or replaced with an updated or new output or loss layer. In at least one embodiment, the initial model 1504 may have previously fine-tuned parameters (e.g., weights and / or biases) retained from previous training, so training or retraining 1514 may not require as much time or processing as training the model from scratch. In at least one embodiment, during model training 1514, when generating predictions on the new customer dataset 1506 by resetting or replacing the output or loss layer of the initial model 1504, the parameters of the new dataset may be updated and readjusted based on the loss calculation associated with the accuracy of the output or loss layer.

[0129] In at least one embodiment, the pre-trained model 1506 may be stored in a data store or registry. In at least one embodiment, the pre-trained model 1506 may have been trained at least partially at one or more facilities other than the facility executing process 1500. In at least one embodiment, to protect the privacy and rights of patients, subjects, or customers at different facilities, the pre-trained model 1506 may have been trained locally using locally generated customer or patient data. In at least one embodiment, the pre-trained model 1506 may be trained using cloud and / or other hardware, but confidential, privacy-protected patient data may not be transferred to, used by, or accessed by any component of the cloud (or other non-local hardware). In at least one embodiment, if the pre-trained model 1506 is trained using patient data from more than one facility, the pre-trained model 1506 may have been trained separately for each facility before training on patient or customer data from another facility. In at least one embodiment, such as when customer or patient data has been published for privacy reasons (e.g., by abandonment, for experimental purposes, etc.), or where customer or patient data is included in a public dataset, customer or patient data from any number of facilities can be used to train a pre-trained model 1506 in and / or off-site, such as in a data center or other cloud computing infrastructure.

[0130] In at least one embodiment, when selecting an application for use in the deployment pipeline, the user may also select a machine learning model for a specific application. In at least one embodiment, the user may not have a model available, so the user may select a pre-trained model to use with the application. In at least one embodiment, the pre-trained model may not be optimized to generate accurate results on the user facility's customer dataset 1506 (e.g., based on patient diversity, demographics, type of medical imaging equipment used, etc.). In at least one embodiment, the pre-trained model may be updated, retrained, and / or fine-tuned for use at various facilities before being deployed into the deployment pipeline for use with one or more applications.

[0131] In at least one embodiment, a user may select a pre-trained model to update, retrain, and / or fine-tune, and this pre-trained model may be referred to as the initial model 1504 of the training system in process 1500. In at least one embodiment, a client dataset 1506 (e.g., imaging data, genomic data, sequencing data, or other data types generated by equipment at the facility) may be used to perform model training (which may include, but is not limited to, transfer learning) on ​​the initial model 1504 to generate a refined model 1512. In at least one embodiment, ground truth data corresponding to the client dataset 1506 may be generated by the training system 1304. In at least one embodiment, ground truth data may be generated at the facility, at least in part, by clinicians, scientists, physicians, or practitioners.

[0132] In at least one embodiment, AI-assisted annotation may be used to generate truth data in some examples. In at least one embodiment, AI-assisted annotation (e.g., implemented using an AI-assisted annotation SDK) may leverage machine learning models (e.g., neural networks) to generate truth data for suggestions or predictions for a customer dataset. In at least one embodiment, a user may use the annotation tool within a user interface (graphical user interface (GUI)) on a computing device.

[0133] In at least one embodiment, user 1510 can interact with the GUI via computing device 1508 to edit or fine-tune (automatic) annotations. In at least one embodiment, polygon editing features can be used to move the vertices of a polygon to more precise or fine-tuned positions.

[0134] In at least one embodiment, once the client dataset 1506 has associated ground truth data, the ground truth data (e.g., from AI-assisted annotations, manual labeling, etc.) can be used to generate a refined model 1512 during model training. In at least one embodiment, the client dataset 1506 can be applied to the initial model 1504 an arbitrary number of times, and the ground truth data can be used to update the parameters of the initial model 1504 until an acceptable level of accuracy is achieved for the refined model 1512. In at least one embodiment, once the refined model 1512 is generated, it can be deployed in one or more deployment pipelines at the facility to perform one or more processing tasks related to medical imaging data.

[0135] In at least one embodiment, the refined model 1512 can be uploaded to a pre-trained model registry for selection by another facility. In at least one embodiment, this process can be performed at any number of facilities, allowing the refined model 1512 to be further refined an arbitrary number of times on a new dataset to generate a more general model.

[0136] Figure 15B This is an example illustration of a client-server architecture 1532 for enhancing annotation tools using a pre-trained annotation model, according to at least one embodiment. In at least one embodiment, an AI-assisted annotation tool 1536 can be instantiated based on the client-server architecture 1532. In at least one embodiment, the AI-assisted annotation tool 1536 in an imaging application can assist radiologists, for example, in identifying organs and abnormalities. In at least one embodiment, the imaging application may include software tools, as a non-limiting example, that help user 1510 identify several extreme points on a specific organ of interest in a raw image 1534 (e.g., in a 3D MRI or CT scan) and receive automatic annotation results for all 2D slices of that specific organ. In at least one embodiment, the results can be stored in a data store as training data 1538 and used as (e.g., but not limited to) ground truth data for training. In at least one embodiment, when computing device 1508 sends extreme points for AI-assisted annotation, for example, a deep learning model can receive this data as input and return inference results for segmenting organs or abnormalities. In at least one embodiment, a pre-instantiated annotation tool (e.g., Figure 15BThe AI-assisted annotation tool 1536 can be enhanced by making API calls (e.g., API call 1544) to a server (such as annotation assistant server 1540), which may include a set of pre-trained models 1542 stored, for example, in an annotation model registry. In at least one embodiment, the annotation model registry may store the pre-trained models 1542 (e.g., machine learning models, such as deep learning models) pre-trained to perform AI-assisted annotation on specific organs or anomalies. These models can be further updated using a training pipeline. In at least one embodiment, the pre-installed annotation tool can be improved over time as new labeled data is added.

[0137] Various embodiments are described by the following terms: 1. At least one processor, said at least one processor comprising: One or more logic units are used for: The machine learning model is used to determine the inference operation to be performed, which includes floating-point values ​​to be exponentially operated on; A fixed-point representation of the value is generated, the fixed-point representation including a bias factor; The fixed-point representation is used to generate an intermediate representation of the value with a defined number of bits; Increment the intermediate representation so that the mantissa of the incremented intermediate representation corresponds to the fixed-point representation; Shift the decimal point of the incremented intermediate representation to produce an exponential approximation of the value; and The inference operation is performed using the machine learning model and the exponential approximation.

[0138] 2. At least one processor as described in Clause 1, wherein the one or more logical units are further configured to: The received value is given an exponential approximation to calculate the normalized probability of the value as the output of the neural network.

[0139] 3. At least one processor as described in Clause 1, wherein the one or more logical units are further configured to: The intermediate representation is incremented by adding the value 1 to it.

[0140] 4. At least one processor as described in Clause 1, wherein the number of decimal point shifts depends in part on the number of bits approximated by the exponent.

[0141] 5. At least one processor as described in Clause 1, wherein the intermediate representation has at most a number of bits determined for the hardware registers to be used to store the intermediate representation.

[0142] 6. At least one processor as described in Clause 1, wherein the one or more logical units are further configured to: After the decimal point is shifted, the sign bit of the incremented intermediate representation is cleared.

[0143] 7. At least one processor as described in Clause 1, wherein the one or more logical units are further configured to: The exponential approximation is provided for use with the SoftMax activation layer of the machine learning model.

[0144] 8. At least one processor as described in Clause 1, wherein the one or more logical units are further configured to: The received value is generated as an exponential approximation without generating an integer representation of the received value.

[0145] 9. At least one processor as described in Clause 1, wherein the at least one processor is included in at least one of the following: A system used to perform simulation operations; A system used to perform simulations to test or validate autonomous machine applications; Systems used to perform digital twin operations; A system for performing optical transmission simulation; A system used for rendering graphics output; A system used to perform deep learning operations; Systems implemented using edge devices; Systems used to generate or present virtual reality (VR) content; A system for generating or presenting augmented reality (AR) content; A system for generating or presenting mixed reality (MR) content; A system containing one or more virtual machines (VMs); A system that is at least partially implemented in a data center; A system for performing hardware tests using simulation; Systems for generating synthetic data; A system for performing generative operations using a large language model LLM; A system for performing generative operations using a visual language model (VLM); A system for performing generative operations using a multimodal language model; A system that uses or deploys one or more inference microservices; A system that includes one or more machine learning models deployed in services or microservices and operating system-level virtualization packages (such as containers); A collaborative content creation platform for 3D assets; or A system that utilizes cloud computing resources at least in part.

[0146] 10. A system comprising: One or more processing units are configured to generate at least one of an exponential or logarithmic approximation using the floating-point representation, in part by generating an intermediate representation using a fixed-point representation of the floating-point representation and a bias factor, adjusting the value of the intermediate representation, and shifting the intermediate representation by a determined number of bits.

[0147] 11. The system as described in Clause 10, wherein at least one of the exponential approximation or logarithmic approximation is used to perform inference operations using a machine learning model.

[0148] 12. The system as described in Clause 11, wherein the one or more processing units are further configured to provide the exponential approximation to calculate the normalized probability of at least one value output from the machine learning model.

[0149] 13. The system as described in Clause 10, wherein the one or more processing units are further configured to increment the intermediate representation such that the mantissa of the intermediate representation represents the fixed-point representation.

[0150] 14. The system as described in Clause 10, wherein the intermediate representation has at most the number of bits determined for the hardware registers to be used to store the intermediate representation.

[0151] 15. The system as described in Clause 10, wherein the one or more processing units are further configured to provide the exponential approximation for use with a SoftMax activation layer of a machine learning model.

[0152] 16. The system as described in Clause 10, wherein the system comprises at least one of the following: A system used to perform simulation operations; A system used to perform simulations to test or validate autonomous machine applications; Systems used to perform digital twin operations; A system for performing optical transmission simulation; A system used for rendering graphics output; A system used to perform deep learning operations; A system for performing generative operations using a large language model LLM; A system for performing generative operations using a visual language model (VLM); A system for performing generative operations using a multimodal language model; Systems implemented using edge devices; Systems used to generate or present virtual reality (VR) content; A system for generating or presenting augmented reality (AR) content; A system for generating or presenting mixed reality (MR) content; A system containing one or more virtual machines (VMs); A system that is at least partially implemented in a data center; A system for performing hardware tests using simulation; Systems for generating synthetic data; A system that uses or deploys one or more inference microservices; A system that includes one or more machine learning models deployed in services or microservices and operating system-level virtualization packages (such as containers); A collaborative content creation platform for 3D assets; or A system that utilizes cloud computing resources at least in part.

[0153] 17. A reasoning system, the reasoning system comprising: A machine learning model, the machine learning model including an activation layer; and One or more processors are used for: Generate a fixed-point representation of the value received by the activation layer in floating-point format, the fixed-point representation including a bias factor; The fixed-point representation is used to generate an intermediate representation of the value with a defined number of bits; Increment the intermediate representation so that the mantissa of the incremented intermediate representation corresponds to the fixed-point representation; Shift the decimal point of the incremented intermediate representation to produce an exponential approximation of the received value; and The probability of the inference result generated by the machine learning model is calculated using the exponential approximation.

[0154] 18. The inference system as described in Clause 17, wherein one or more of the probabilities are calculated from the received values ​​without conversion to single-precision format or integer values.

[0155] 19. The inference system as described in Clause 17, wherein the probability is a normalized probability calculated using an exponential approximation of a plurality of values ​​received from the activation layer.

[0156] 20. The reasoning system as described in Clause 17, wherein the reasoning system comprises at least one of the following: A system used to perform simulation operations; A system used to perform simulations to test or validate autonomous machine applications; Systems used to perform digital twin operations; A system for performing optical transmission simulation; A system used for rendering graphics output; A system used to perform deep learning operations; A system for performing generative operations using a large language model LLM; A system for performing generative operations using a visual language model (VLM); A system for performing generative operations using a multimodal language model; Systems implemented using edge devices; Systems used to generate or present virtual reality (VR) content; A system for generating or presenting augmented reality (AR) content; A system for generating or presenting mixed reality (MR) content; A system containing one or more virtual machines (VMs); A system that is at least partially implemented in a data center; A system for performing hardware tests using simulation; Systems for generating synthetic data; A system that uses or deploys one or more inference microservices; A system that includes one or more machine learning models deployed in services or microservices and operating system-level virtualization packages (such as containers); A collaborative content creation platform for 3D assets; or A system that utilizes cloud computing resources at least in part.

[0157] Other variations are within the spirit of this disclosure. Therefore, although the disclosed technology is readily adaptable to various modifications and alternative constructions, certain embodiments thereof are illustrated in the accompanying drawings and have been described in detail above. However, it should be understood that the disclosure is not intended to be limited to one or more specific forms disclosed, but rather, it is intended to cover all modifications, alternative constructions, and equivalents falling within the spirit and scope of this disclosure as defined in the appended claims.

[0158] Unless otherwise stated or obviously contradicted by the context, the terms “a,” “an,” and “the,” and similar references, used in the context of describing the disclosed embodiments (particularly in the context of the appended claims), should be interpreted as encompassing both singular and plural forms, rather than as definitions of the terms. Unless otherwise stated, the terms “comprising,” “having,” “including,” and “containing” should be interpreted as open-ended terms (meaning “including, but not limited to”). The term “connection” (referring to a physical connection where not modified) should be interpreted as partially or wholly contained, attached to, or joined together, even with some intervention. Unless otherwise indicated herein, references to numerical ranges herein are intended only as a way of abbreviating each individual value falling within that range, and each individual value is incorporated into the specification as if it were separately described herein. Unless otherwise indicated or contradicted by the context, the use of the terms “set” (e.g., “item set”) or “subset” should be interpreted as a non-empty set comprising one or more members. Furthermore, unless otherwise indicated or contradicted by the context, the term “subset” of the corresponding set does not necessarily mean an appropriate subset of the corresponding set, but rather that the subset and the corresponding set can be equal.

[0159] Unless otherwise explicitly stated or clearly contradicted by the context, connective phrases such as “at least one of A, B, and C” or “at least one of A, B, and C” are understood in the context to generally refer to items, terms, etc., which can be A or B or C, or any non-empty subset of the set A, B, and C. For example, in an illustrative example of a set with three members, the connective phrases “at least one of A, B, and C” and “at least one of A, B, and C” refer to any of the following sets: {A}, {B}, {C}, {A, B}, {A, C}, {B, C}, {A, B, C}. Therefore, such connective language is generally not intended to imply that some embodiments require the presence of at least one of A, at least one of B, and at least one of C. Additionally, unless otherwise stated or contradicted by the context, the term “multiple” indicates a plural state (e.g., “multiple items” means multiple items). Multiple means at least two items, but more may be indicated if explicitly stated or by the context. Furthermore, unless otherwise stated or clearly understood from the context, the phrase “based on” means “at least partially based on” rather than “based on only”.

[0160] Unless otherwise stated herein or clearly contradicted by the context, the operations of the processes described herein may be performed in any suitable order. In at least one embodiment, processes such as those described herein (or variations thereof and / or combinations thereof) are executed under the control of one or more computer systems configured with executable instructions and are implemented as code (e.g., executable instructions, one or more computer programs, or one or more applications) that is executed jointly on one or more processors via hardware or a combination thereof. In at least one embodiment, the code is stored on a computer-readable storage medium, for example, in the form of a computer program comprising a plurality of instructions executable by one or more processors. In at least one embodiment, the computer-readable storage medium is a non-transitory computer-readable storage medium that excludes transient signals (e.g., propagating transient electrical or electromagnetic transmissions) but includes non-transitory data storage circuitry (e.g., buffers, caches, and queues). In at least one embodiment, code (e.g., executable code or source code) is stored on one or more non-transitory computer-readable storage media (or other memory for storing executable instructions) on which executable instructions are stored, which, when executed by one or more processors of a computer system (i.e., as a result of execution), cause the computer system to perform the operations described herein. In at least one embodiment, the set of non-transitory computer-readable storage media comprises multiple non-transitory computer-readable storage media, and one or more of the individual non-transitory storage media lack all the code, but the multiple non-transitory computer-readable storage media collectively store all the code. In at least one embodiment, the executable instructions are executed such that different instructions are executed by different processors; for example, the non-transitory computer-readable storage media store the instructions, and the main central processing unit (“CPU”) executes some instructions while the graphics processing unit (“GPU”) executes other instructions. In at least one embodiment, different components of the computer system have separate processors, and the different processors execute different subsets of the instructions.

[0161] Therefore, in at least one embodiment, the computer system is configured to implement one or more services that perform the processes described herein individually or collectively, and such a computer system is configured with suitable hardware and / or software that allows the implementation of the operations. Furthermore, the computer system implementing at least one embodiment of this disclosure is a single device, and in another embodiment it is a distributed computer system comprising multiple devices operating in different ways, such that the distributed computer system performs the operations described herein, and that a single device does not perform all the operations.

[0162] The use of any and all examples or exemplary language (e.g., "such as") provided herein is intended only to better illustrate embodiments of this disclosure and does not constitute a limitation on the scope of the disclosure unless otherwise required. No language in the specification should be construed as indicating that any unclaimed element is essential to the practice of the disclosure.

[0163] All references cited in this article, including publications, patent applications and patents, are incorporated herein by reference to the same extent that each reference is individually and specifically indicated to be incorporated herein by reference and that its entire contents are set forth herein.

[0164] The terms “coupled” and “connected”, and their derivatives, may be used in the specification and claims. It should be understood that these terms may not be intended to be synonyms with each other. Rather, in certain examples, “connected” or “coupled” may be used to indicate that two or more elements are in direct or indirect physical or electrical contact with each other. “Coupled” may also mean that two or more elements are not in direct contact with each other, but still cooperate or interact with each other.

[0165] Unless otherwise expressly stated, it will be understood that throughout this specification, terms such as “processing,” “computing,” “determining,” etc., refer to the actions and / or processes of a computer or computing system or similar electronic computing device that process and / or convert data represented as physical quantities (e.g., electrons) in the registers and / or memory of the computing system into other data represented as physical quantities in the memory, registers, or other such information storage, transmission, or display devices of the computing system.

[0166] In a similar manner, the term "processor" can refer to any device or part of memory that processes electronic data from registers and / or memory and converts that electronic data into other electronic data that can be stored in registers and / or memory. As a non-limiting example, a "processor" can be a CPU or a GPU. A "computing platform" can include one or more processors. As used herein, a "software" process can include, for example, software and / or hardware entities that perform work over time, such as tasks, threads, and intelligent agents. Similarly, each process can refer to multiple processes that execute instructions sequentially or intermittently, sequentially, or in parallel. The terms "system" and "method" are used interchangeably herein, provided that a system can embody one or more methods, and a method can be considered a system.

[0167] This document refers to the process of acquiring, obtaining, receiving, or inputting analog or digital data into a subsystem, computer system, or computer-implemented machine. Acquiring, obtaining, receiving, or inputting analog and digital data can be accomplished in various ways, such as by receiving data as a parameter to a function call or an application programming interface (API) call. In some implementations, the process of acquiring, obtaining, receiving, or inputting analog or digital data can be accomplished by transmitting data via a serial or parallel interface. In another implementation, the process of acquiring, obtaining, receiving, or inputting analog or digital data can be accomplished by transmitting data from a providing entity to an acquiring entity via a computer network. Reference can also be made to providing, outputting, transmitting, sending, or presenting analog or digital data. In various examples, the process of providing, outputting, transmitting, sending, or presenting analog or digital data can be implemented by transmitting data as an input or output parameter to a function call, an API call, or an inter-process communication mechanism.

[0168] While the discussion above illustrates example implementations of the described technologies, other architectures can be used to implement the described functionality and are intended to fall within the scope of this disclosure. Furthermore, although specific assignments of responsibilities have been defined above for discussion purposes, various functions and responsibilities can be assigned and divided in different ways depending on the circumstances.

[0169] Furthermore, although the subject matter has been described in language specific to structural features and / or methodological actions, it should be understood that the subject matter claimed in the appended claims is not necessarily limited to the specific features or actions described. Rather, specific features and actions are disclosed as exemplary forms for implementing the claims.

Claims

1. At least one processor, said at least one processor comprising: One or more logic units are used for: The machine learning model is used to determine the inference operation to be performed, which includes floating-point values ​​to be exponentially operated on; A fixed-point representation of the value is generated, the fixed-point representation including a bias factor; The fixed-point representation is used to generate an intermediate representation of the value with a defined number of bits; Increment the intermediate representation so that the mantissa of the incremented intermediate representation corresponds to the fixed-point representation; The decimal point of the incremented intermediate representation is shifted to produce an exponential approximation of the value; as well as The inference operation is performed using the machine learning model and the exponential approximation.

2. The at least one processor as claimed in claim 1, wherein, The one or more logic units are also used for: The received value is provided as an exponential approximation to calculate the normalized probability of the value as the output of the neural network.

3. The at least one processor as claimed in claim 1, wherein, The one or more logic units are also used for: The intermediate representation is incremented by adding the value 1 to it.

4. The at least one processor as claimed in claim 1, wherein, The number of decimal places shifted depends in part on the number of places approximated by the exponent.

5. The at least one processor as claimed in claim 1, wherein, The intermediate representation has at most the number of bits determined for the hardware registers to be used to store the intermediate representation.

6. The at least one processor as claimed in claim 1, wherein, The one or more logic units are also used for: After the decimal point is shifted, the sign bit of the incremented intermediate representation is cleared.

7. The at least one processor as claimed in claim 1, wherein, The one or more logic units are also used for: The exponential approximation is provided for use with the SoftMax activation layer of the machine learning model.

8. The at least one processor as claimed in claim 1, wherein, The one or more logic units are also used for: The received value is generated as an exponential approximation without generating an integer representation of the received value.

9. The at least one processor as claimed in claim 1, wherein, The at least one processor is included in at least one of the following: A system used to perform simulation operations; A system used to perform simulations to test or validate autonomous machine applications; Systems used to perform digital twin operations; A system for performing optical transmission simulation; A system used for rendering graphics output; A system used to perform deep learning operations; Systems implemented using edge devices; Systems used to generate or present virtual reality (VR) content; A system for generating or presenting augmented reality (AR) content; A system for generating or presenting mixed reality (MR) content; A system containing one or more virtual machines (VMs); A system that is at least partially implemented in a data center; A system for performing hardware tests using simulation; Systems for generating synthetic data; A system for performing generative operations using a large language model LLM; A system for performing generative operations using a visual language model (VLM); A system for performing generative operations using a multimodal language model; A system that uses or deploys one or more inference microservices; A system that includes one or more machine learning models deployed in services or microservices and operating system-level virtualization packages (such as containers); A collaborative content creation platform for 3D assets; or A system that utilizes cloud computing resources at least in part.

10. A system comprising: One or more processing units are configured to generate at least one of an exponential or logarithmic approximation using the floating-point representation, in part by generating an intermediate representation using a fixed-point representation of the floating-point representation and a bias factor, adjusting the value of the intermediate representation, and shifting the intermediate representation by a determined number of bits.

11. The system of claim 10, wherein, At least one of the exponential or logarithmic approximations will be used to perform inference operations using a machine learning model.

12. The system of claim 11, wherein, The one or more processing units are also configured to provide the exponential approximation to calculate the normalized probability of at least one value output from the machine learning model.

13. The system of claim 10, wherein, The one or more processing units are further configured to increment the intermediate representation such that the mantissa of the intermediate representation represents the fixed-point representation.

14. The system of claim 10, wherein, The intermediate representation has at most the number of bits determined for the hardware registers to be used to store the intermediate representation.

15. The system of claim 10, wherein, The one or more processing units are also configured to provide the exponential approximation for use with the SoftMax activation layer of the machine learning model.

16. The system of claim 10, wherein, The system includes at least one of the following: A system used to perform simulation operations; A system used to perform simulations to test or validate autonomous machine applications; Systems used to perform digital twin operations; A system for performing optical transmission simulation; A system used for rendering graphics output; A system used to perform deep learning operations; A system for performing generative operations using a large language model LLM; A system for performing generative operations using a visual language model (VLM); A system for performing generative operations using a multimodal language model; Systems implemented using edge devices; Systems used to generate or present virtual reality (VR) content; A system for generating or presenting augmented reality (AR) content; A system for generating or presenting mixed reality (MR) content; A system containing one or more virtual machines (VMs); A system that is at least partially implemented in a data center; A system for performing hardware tests using simulation; Systems for generating synthetic data; A system that uses or deploys one or more inference microservices; A system that includes one or more machine learning models deployed in services or microservices and operating system-level virtualization packages (such as containers); A collaborative content creation platform for 3D assets; or A system that utilizes cloud computing resources at least in part.

17. A reasoning system, the reasoning system comprising: A machine learning model, wherein the machine learning model includes an activation layer; as well as One or more processors are used for: Generate a fixed-point representation of the value received by the activation layer in floating-point format, the fixed-point representation including a bias factor; The fixed-point representation is used to generate an intermediate representation of the value with a defined number of bits; Increment the intermediate representation so that the mantissa of the incremented intermediate representation corresponds to the fixed-point representation; The decimal point of the incremented intermediate representation is shifted to produce an exponential approximation of the received value; as well as The probability of the inference result generated by the machine learning model is calculated using the exponential approximation.

18. The reasoning system of claim 17, wherein, One or more of the probabilities are calculated from the received values ​​without converting them to single-precision format or integer values.

19. The reasoning system of claim 17, wherein, The probability is a normalized probability calculated using an exponential approximation of multiple values ​​received from the activation layer.

20. The reasoning system of claim 17, wherein, The reasoning system includes at least one of the following: A system used to perform simulation operations; A system used to perform simulations to test or validate autonomous machine applications; Systems used to perform digital twin operations; A system for performing optical transmission simulation; A system used for rendering graphics output; A system used to perform deep learning operations; A system for performing generative operations using a large language model LLM; A system for performing generative operations using a visual language model (VLM); A system for performing generative operations using a multimodal language model; Systems implemented using edge devices; Systems used to generate or present virtual reality (VR) content; A system for generating or presenting augmented reality (AR) content; A system for generating or presenting mixed reality (MR) content; A system containing one or more virtual machines (VMs); A system that is at least partially implemented in a data center; A system for performing hardware tests using simulation; Systems for generating synthetic data; A system that uses or deploys one or more inference microservices; A system that includes one or more machine learning models deployed in services or microservices and operating system-level virtualization packages (such as containers); A collaborative content creation platform for 3D assets; or A system that utilizes cloud computing resources at least in part.

Images