586 results about "Improved performance" patented technology

Embodiments provide modular and reconfigurable battery energy storage systems (RBESSs) exhibiting improved performance and power management capabilities. A power management system is provided to determine a power configuration for a plurality of power storage modules, each having a power storage device (e.g., a battery). The optimized power configuration balances a power load across active power storage modules (e.g., power storage devices being charged or discharged). The power management system is configured to control each power storage module / device on an individual basis, thereby enabling a greater degree of control and flexibility to achieve optimal balancing across a plurality of separate power storage devices within a power system (e.g., a battery pack).

This disclosure describes systems, methods, and devices related to enhanced memory integration. The device may include a compute chiplet configured as a System-on-a-Chip (SoC). The device may include a logic die circuitry coupled to the compute chiplet through a high-speed link. The device may include a memory interface that connects the logic die circuitry to on-package memory. The device may include control circuitry within the logic die circuitry configured to treat the on-package memory as a memory-side cache for an off-package memory. The device may dynamically migrate memory pages between the on-package memory and the off-package memory based on memory access patterns. The device may facilitate efficient data management, optimize memory utilization, and support scalable memory architectures for improved performance.

A first neural network (NN) model may generate labels for training a second NN model. The second NN model may represent instances of a NN model operating on multiple different devices (e.g., decentralized user and / or edge devices). The system may include using a “teacher” model to process data received by one or more of the devices to generate a labeled dataset. The system may use the labeled dataset and a “student” model to calculate gradient data for updating the student model. The student model may be the same or similar to NN model instances operating on the devices. The system may validate the updated student model to determine, for example, whether it exhibits improved performance when processing the newly received data and / or historical data. The system may distribute the validated update to the devices.

A heat pipe includes an extruded profile body, with a hollow body closed at the ends, filled with a predefined volume of diphasic working fluid. A plurality of longitudinal channels are included, with each having a section delimited by a bottom formed by one tubular peripheral wall of the profiled body, and laterally by two longitudinal dividers. A circumferential transfer channel, which is arranged transversely to the local axial direction and provides mutual fluid connection between the longitudinal channels, is provided at a position along the longitudinal path and or at an end, where the longitudinal dividers are interrupted, partially or completely, in the area of the circumferential channel, with optional use of a closure ring.

The invention provides a chip power consumption control method and device and a storage medium, and relates to the technical field of chip power consumption control, and the method comprises the steps: obtaining a target power consumption value, a current power consumption value and a current frequency value; determining a power consumption difference proportion value according to the current power consumption value and a target power consumption value; according to the current frequency value and the power consumption difference proportion value, obtaining a frequency increase and decrease value which is in positive correlation with the power consumption difference proportion value; determining a target frequency value according to the frequency increase / decrease value and the current frequency value; and adjusting the working frequency and the working voltage according to the target frequency value. The target frequency value meeting the target power consumption value condition can be quickly determined without step-by-step frequency increasing and decreasing, the frequency modulation efficiency can be improved, the performance and the temperature control performance can be improved, the working power consumption can be controlled to be close to the target power consumption value, and the power consumption requirement can be met. And meanwhile, smaller-granularity frequency modulation is realized, and the frequency for giving full play to the performance of the chip is obtained under the limitation of power consumption, so that the utilization efficiency of the performance of the chip is improved.

A lens form for a low parallax imaging device includes a plurality of imaging lens elements arranged to capture and image polygonal fields of view. The lens form includes a compression lens element group and a wide-angle lens element group.

A method for manufacturing a semiconductor device is disclosed. Methods involve forming a stack of alternating semiconductor channels and interposers on a substrate with sacrificial structures between the interposers. Source / drain openings are formed, and strain in the channel is modified. Source / drain structures are formed in the openings, and a dielectric layer is deposited. The resulting device features a stacked nanostructure having internal spacers of varying height, enabling improved performance in electronic devices. The embodiment of the invention also relates to a semiconductor device and a forming method thereof.

A system and related method, including system memory with a source memory block and a destination memory block each of a memory block size, a first volatile memory and a second volatile memory, each of a volatile memory size. The system includes processing circuitry to receive a garbage collection request associated with the destination memory block. The processing circuitry then determines whether the memory block size is greater than the volatile memory size and pauses data change operations but allows for passive operations to continue. While data change operations are paused the processing circuitry loads a first portion of data of the source memory block to the first volatile memory and a second portion of the data to the second volatile memory. The processing circuitry writes each portion of data from the volatile memory to the destination memory block to complete garbage collection and unpauses the data change operations.

A computer-implemented method of configuring a communication network. Devices of the network are configured to communicate according to scheduling parameters that indicate a timing of transmitting and / or receiving data over the communication network. To configure the network, the method uses a configuration procedure that is customizable by a set of customization parameters. For a current scheduling instance, representing multiple data streams to be scheduled, the configuration procedure is applied to determine current scheduling parameters, which are provided to the respective devices. Concurrently with the current customization parameters being used, the customization parameters are optimized with respect to a performance indicator of scheduling the current scheduling instance according to the customization parameters, with the updated customization being used to determine scheduling parameters for future scheduling instances with improved performance with respect to the performance indicator.

A flyback converter comprising: a transformer having a primary side winding and a secondary side winding; a primary side switch located at the primary side of the transformer; two secondary side switches located at the secondary side of the transformer; in which the two secondary side switches are connected in parallel and have different characteristics; and a control unit configured to generate a rectification signal and a Zero Voltage Switching (ZVS) pulse and to drive the two secondary side switches; in which a secondary side switch is driven by the rectification signal and the other secondary side switch is driven by the ZVS pulse. At the end of a switching cycle, before turning on the primary side switch: the control unit is configured to generate the ZVS pulse in the secondary side winding such that the parasitic capacitor of the primary side switch is discharged, and to consequently turn on the primary side switch in ZVS conditions or near ZVS conditions.

A golf club head and golf club having large head dimensions, and large face characteristics, as well as unique mass property attributes driven by relationships not often considered in conventional club head design, to achieve a resistance to squaring the club head during a golf swing that is comfortable to the novice golfer, while increasing stability during off-center impacts and obtaining preferred launch characteristics. This is achieved in part via establishing a club head configuration and associated weight distribution to yield a center of gravity location that results in a preferred magnitude of Delta1 and Delta2 values, CG angle, moments of inertia, and associated ratios, relationships, and club head mass property characteristics influenced by these variables to achieve improved performance.

A method of serving a generative transformer model includes determining a batch size to use in processing inference requests and allocating at least one prompt pipeline and at least on token pipeline to the generative transformer model to process the batch of inference requests. The number of prompt pipelines and the number of token pipelines, and the depths of the pipelines are determined based on the batch size, an average prompt length, a cache requirement per stage, and a memory footprint of model weights for the generative model per stage using a resource allocator component of the model serving system. Cache streaming is used to stream prompt cache from prompt pipelines to token pipelines to generate tokens. Cache streaming involves gather-copy operations which may be performed using compute kernels.

The invention discloses an irregular texture product appearance defect detection method and device, and the method comprises the steps: collecting N good product sample images through an area-array camera, taking one image as a reference image, and obtaining a product contour shape positioning template through an edge positioning extraction function; preprocessing all good product sample images by using a detection preprocessing function, obtaining position information of good product samples through a product contour shape positioning template, aligning all positions of the good product samples by using an affine transformation function, and introducing an improved performance function for training to obtain a good product detection model; collecting a detection image, performing positioning preprocessing, and obtaining position information of a target by using a product contour shape positioning template; carrying out detection preprocessing, transforming the preprocessed image and a good product detection model to the same reference position, and carrying out point-by-point comparison to screen out different parts; according to the invention, rapid and accurate detection of the appearance defects of the irregular texture detection area is realized under the conditions that the number of die-cutting product defect samples is small and the variety of products is large.

The invention provides an exhaust system capable of achieving both improved performance and a damping effect when an exhaust valve is open. An exhaust system (30) comprises an exhaust muffler (34) with an expansion chamber (A), an exhaust outlet pipe (55) for expelling combustion gas (G) from the exhaust muffler (34), and an exhaust valve mechanism (40) for adjusting the flow rate of the combustion gas (G) with an on / off exhaust valve (42). The exhaust outlet pipe (55) is connected to the expansion chamber (A). The exhaust valve (42) is located in an upstream exhaust pipe (50) for introducing the combustion gas (G) into the expansion chamber (A). The combustion gas (G) is also introduced into the expansion chamber (A) when the exhaust valve (42) is closed.

There is provided a system and method for examining a semiconductor specimen. The method includes obtaining a runtime image of a semiconductor specimen acquired by an examination tool; processing the runtime image to create one or more image strips each containing an edge, and for each image strip, extracting a sequence of topo points representative of a contour of the edge therein; providing the sequence of topo points for each image strip to a trained machine learning (ML) model to be processed, and obtaining, as an output of the ML model, a sequence of updated topo points; and obtaining measurement data on the runtime image using the sequence of updated topo points, wherein the measurement data has improved performance with respect to at least one metrology metric.

A torque control fastener system having a fastener nut, a helical wire insert, and a shaft, where the nut and helical wire insert have unique hardness and coefficient of thermal expansion relationships that produce improved performance.

The invention relates to a brake pad or a disc brake caliper, comprising a support plate receiving a friction material. According to the invention, the support plate is made from low-alloy carbon steel containing less than 0.3% by weight of carbon and vanadium in an amount between 0.2 and 0.3% by weight.

A system is disclosed, including a display, a processor and a memory. The memory stores instructions that, when executed by the processor, configure the system to perform operations. Active region data is generated that includes, for each of one or more active regions, active region location data and active region content. The active region data is transmitted to a display having a display area. For each active region, the active region content is displayed at an active region location of the display area based on the active region location data, the active region content being displayed at a higher spatiotemporal information density than content displayed in the display area outside of the active regions.

The present technology, roughly described, provides for mocking an application program interface (API) using a large language model (LLM). The present system generates a prompt with API signature information and API desired behavior information. The prompt can include instructions, library functions, examples, executed programs, and a current function invocation, as well as other content. The prompt can be generated, and submitted to an LLM to mock an API and generate a response. The response can be audited and the LLM can be fine tuned to provide improved performance in subsequent calls.

Provided is a computer program, an information processing device, and a method that are used to estimate a health state of a target user based on a history of behaviors executed by the target user, and have at least partially improved performance. According to one embodiment, a computer program can cause, by being executed at least one processor, the at least one processor to function to: acquire target behavior data for identifying a history of behavior executed by a target user; and output, from an estimation model generated by executing supervised learning, target sleep state data for identifying a sleep state of the target user, the target sleep state data including at least one of target property data for identifying a sleep property of the target user, target rhythm data for identifying a sleep rhythm of the target user, target time data for identifying a sleep time of the target user, and target category data for identifying a sleep category to which the target user belongs, by inputting the target behavior data to the estimation model.

An electronic device includes a multilevel package substrate, a semiconductor die, and a package structure, the multilevel package substrate has first, second, and third levels including respective dielectric layers and conductive features, the first level including a first trace layer with an antenna and a first via layer with a portion of a ground wall laterally spaced outward from and surrounding the antenna, and the second level including a second trace layer having a ground plane connected to the ground wall, the semiconductor die attached to the first level of the multilevel package substrate, and the package structure including a molding compound enclosing the semiconductor die and extending on a side of the antenna, where the package structure mold compound maters and thickness can be tuned for improved performance.

This disclosure provides systems, methods, and devices for control and management of production home builders. According to aspects, computerized tools described herein may enable home builders to react more quickly and consistently to new data relating to market forces and home buyer demand, enabling the home builder to improve home production and its performance. Some disclosed features may enable and provide improved performance metric tracking or home demand classification, including classifying homes and home builders based on a pacing metric. In some aspects, a system may determine target performance paces for the home builder based on received home status data, such as a completion status related to the construction level of the homes. As such, the disclosed pacing tool may enable home builders to more accurately match the home transactions pace to the home production pace, which may improve marginal gains and provide for greater influence over the housing market.

This disclosure provides systems, methods, and devices for control and management of production home builders. According to aspects of the disclosure, computerized tools described herein may enable home builders to react more quickly and consistently to new data relating to market forces and home buyer demand. Some features disclosed herein may enable and provide improved performance metric tracking, including dynamically determining a pacing metric based on data. In some aspects, the system may determine a performance pace for a first plurality of homes based on a comparison of home status data for the first plurality of homes to performance data for a second plurality of homes. Such operations may beneficially enable effective pacing of home builder performance that more accurately matches the home sales pace to the home production pace. Additionally, the system may output instructions of actionable steps to improve home production and performance in the market.

Techniques to improve performance of loading data from memory addresses implemented in a storage capacity of a memory sub-system. A connection from the memory sub-system to a host system supports both a protocol for cache-coherent memory access to a memory device implemented in the storage capacity and a protocol for storage access. A cache memory is used in the memory sub-system to cache pages of the memory device for accessing over the connection. The host system can enter hints about future memory accesses in a storage access queue, which can be configured in the memory device, or a memory of the host system. Based on the hints the memory sub-system can prefetch pages from the storage capacity into the cache memory for improved performance in servicing requests from the host system to load data from the memory device.

Some example embodiments provide an optical proximity correction (OPC) method using an OPC model having improved performance and / or a method of manufacturing a mask by using the OPC method. An OPC method includes receiving a design layout of a target pattern, generating a first OPC model on the design layout, in which an optical effect of an exposure process is reflected, generating a second OPC model in which a characteristic of a photoresist in the exposure process is reflected, and performing a simulation using the first and second OPC models to obtain an OPC-performed design layout. The generating the second OPC model includes differently applying a combination of kernel functions, used in the second OPC model, to each pattern region.

Rather than using low power mode since there is no indication of a sleep mode type to the data storage device, the sleep mode type is inferred by the data storage device, or supplied by the host. In so doing, the sleep type communication is improved. After the system returns power to the data storage device, there may be a read workload, depending on the sleep type. The workload is characterized by a low-queue depth (QD) sequential read from a specific area of the storage medium that was written to just prior to shut down. In response to inference or host cue, the data storage device will modify an operation so that data storage device is optimized to the sleep mode type, resulting in improved performance and power consumption.

Embodiments of the present disclosure relate to a communication device, method, and apparatus, and a computer-readable medium. In one aspect, a first access point (AP) device assigns a device identification to a station (STA) device during performing an association process with the STA device. In addition, the first AP device transmits the device identification assigned to the STA device to at least one of a second AP device or a control device. In this way, the performance of a communication network can be improved. For example, an EasyMesh network with improved performance can be achieved.

Described are systems and methods for client side enrichment and transform via dynamic logic for analytics across various platforms for improved performance, features, and uses. Analytics data collected in client applications is transformed and enriched before being sent to the downstream pipeline using native code and logic bundled into the core application code. The additional logic specific to manipulation of analytics may be unbundled from client-side application code and still be executed on on-device to achieve the same result. The logic may be written in a single language, such as JavaScript, and run across all clients including web browser and mobile operating systems.

A radio frequency package for an electronic device with a reduced size and improved performance is presented herein. The radio frequency package includes a front-end package, transceiver dies and at least one antenna array providing a wireless communication functionality for the radio frequency package. The front-end package includes a radio frequency front-end die and at least one power control die connected to an insulation film substrate via first connectors. Each of the transceiver dies is connected to the at least one antenna array using second connectors.