193 results about "Register bank" patented technology

The invention discloses an attention mechanism calculation method and device, a storage medium and a product, and the method comprises the steps: carrying out matrix multiplication operation through employing a query matrix block of a first register block and a key matrix block of a shared memory, obtaining a first product matrix block, and writing the first product matrix block into a second register block; performing exponential operation by using the first product matrix blocks to obtain sub-matrix blocks, writing the sub-matrix blocks into a second register block in a covering manner, and writing the sub-matrix blocks into a third register block in the form of a target precision type; performing matrix multiplication operation by using the sub-matrix blocks of the third register block and the value matrix blocks of the shared memory to obtain second product matrix blocks, and writing the second product matrix blocks into a second register block; performing softmax operation by using the second product matrix blocks to obtain attention result matrix blocks, and writing the attention result matrix blocks into a fourth register block; and writing the attention result matrix of the fourth register group into the shared memory in blocks. According to the embodiment of the invention, overflow of the register can be avoided, and the utilization of hardware resources is maximized.

The invention discloses an attention calculation implementation method and device, a medium, equipment and a product, and the method comprises the steps: loading the current to-be-calculated ith query block from a shared memory to a first register group of a consumer thread group, and carrying out the internal splitting calculation of the query block and a key block, so as to obtain corresponding attention score blocks; sequentially obtaining MK attention score blocks, and executing attention fusion calculation of mixing precision with the corresponding value blocks to obtain an attention output block corresponding to the ith query block; and after the attention output blocks are logically divided into N2 batches, a specified register group for storing target precision type data in the fusion calculation process is multiplexed and executed according to batches, target precision type conversion is carried out, and an output result after conversion of each batch is written back to a shared memory. According to the method, the existing hardware resources can be efficiently utilized to improve the calculation performance, and the method is particularly suitable for large-size query block and key block scenes.

The invention discloses an embedded SoC-level real-time monitoring and analyzing device. The embedded SoC-level real-time monitoring and analyzing device is composed of a bus comparator, a system event counter and a control register set. The bus comparator collects access signals of an address bus, a data bus and an instruction bus in real time, compares the access signals with a preset matching rule, generates a trigger signal for driving the event state register to record and output an interrupt signal or a debugging signal during matching, and takes the trigger signal as an input event of the system event counter; the system event counter monitors various events generated in the operation process, and when the monitoring result is equal to a preset reference value, a trigger signal is generated; and the control register group configures the matching condition of the bus comparator, sets the counting mode of the system event counter, and reads the comparison state, the event counting result and related monitoring information through the bus interface. According to the method, a configurable event capture mechanism and multiple counting modes are introduced in a hardware level, so that the dependence on software debugging is reduced, and the real-time performance of monitoring and analysis is improved.

The invention discloses an attention mechanism calculation method and device, a storage medium and a product, and the method comprises the steps: carrying out the matrix multiplication operation of a query matrix block and a key matrix block, obtaining a first product matrix block of a first precision type, writing the first product matrix block into a second register group, and then carrying out the index operation, and obtaining a molecular matrix block; writing the quantized sub-matrix blocks into a third register block according to a second precision type; writing the sub-matrix blocks into a fourth register block according to a third precision type, and performing layout conversion through a shared memory; performing softmax operation by using the molecular matrix blocks and the value matrix blocks after layout conversion to obtain attention result matrix blocks of a first precision type, and writing the attention result matrix blocks into a fifth register block; and writing the attention result matrix blocks into the shared memory according to the third precision type. The embodiment of the invention can reduce the accumulative error of the intermediate operation result.

The invention provides a memory access system oriented to vector and matrix calculation, the system comprises a vector memory access unit, a matrix memory access unit, a vector register group and a matrix register group, the vector memory access unit is connected with a memory interface and the vector register group, reads elements of a one-dimensional data structure or a two-dimensional data structure from a memory, and stores the elements of the one-dimensional data structure or the two-dimensional data structure; vector data are generated through data reorganization operation, a matrix access unit is connected with a memory interface and a matrix register set, matrix block data of a two-dimensional data structure are read from a memory together with the matrix access unit, matrix data are generated after data reorganization, and the matrix data are broadcasted to one or more computing units according to rows or columns. And performing calculation on the matrix data and the vector data. In the memory access system, the vector memory access unit and the matrix memory access unit can load data in parallel, and the utilization rate is improved through a plurality of computing units, so that the problems of low memory access efficiency and low memory bandwidth utilization rate are solved.

The invention discloses a data processing device, a data processing method, a chip and electronic equipment, and relates to the technical field of data processing. The data processing device comprises a storage module which is configured to store x groups of original data vectors, the original data vectors comprise m data elements, and x and m are positive integers; the calculation module is configured to access the original data vectors stored in the storage module, and reconstruct the x groups of original data vectors into one or more intermediate matrixes according to the parallel processing capability of the single-instruction multi-data execution unit and a target value k of a Top-k operator; loading the intermediate matrix into a vector register block according to columns, carrying out odd-even merging sorting by utilizing a single-instruction multi-data execution unit, and realizing in-line sorting on the intermediate matrix; and loading the intermediate matrix subjected to in-line sorting into a vector register block according to columns, and performing one or more rounds of merging sorting by utilizing a single-instruction multi-data execution unit to obtain top-k data elements in x groups of original data vectors.

The invention provides an AXI bus monitoring system and method, the system comprises a host side monitor, a slave side monitor and an error summarization processing module, and the error summarization processing module is respectively connected with the host side monitor and the slave side monitor; a plurality of register groups are arranged in each monitor, and each register group corresponds to one timer and one state machine and is used for monitoring one read or write transaction; the timer is used for carrying out timeout counting on the read / write transaction, and when the count exceeds a preset value, a timeout error is triggered to be reported to the error summarization processing module; the state machine is used for tracking which stage the read or write transaction is performed in real time; the register block is used for latching an AXI bus signal so that a monitor constructs data to complete access; and the error summarization processing module is used for executing corresponding processing according to the error codes reported by the host side monitor and / or the slave side monitor and the information latched by the register block. The system can effectively avoid system deadlock.

The invention provides a matrix multiplication operation method and device and a matrix transposition operation method and device of a processor. The matrix multiplication operation method of the processor comprises the following steps: taking a first matrix of which the storage sequence is inconsistent with the loading sequence for carrying out element-by-element multiplication operation as a second matrix, loading the second matrix through a register set in the processor, and dividing the second matrix into a plurality of third matrixes, replacing elements in a plurality of nth register groups included in the register set; for each nth register group, reading elements in the nth register group according to an alternating sequence of the elements of the first register and the elements of the second register through the (n + 1) th register group; when iteration is carried out for n to N-1, the obtained sixth matrix is stored in the memory; and performing element-by-element multiplication and accumulation on the elements at each position of the first matrix and the sixth matrix. According to the invention, the time required by matrix multiplication operation can be reduced.

The invention provides an RISC-V tensor instruction set extension method based on an intelligent processor. The method comprises the steps that a special tensor configuration register set is introduced to store tensor element information and operation element information; encoding a tensor extension instruction by adopting a 64-bit fixed-length instruction encoding format; and setting high-capacity on-chip tight coupling storage with uniform addressing in a processor core, and taking the high-capacity on-chip tight coupling storage as a direct operation space of all tensor expansion instructions. The invention further provides an intelligent processor-based RISC-V tensor instruction set extension device, a storage medium and electronic equipment. According to the method, the RISC-V tensor instruction set extension which supports high-dimensional tensor operation, is dense in calculation and is friendly to hardware can be realized.

The invention provides an artificial intelligence chip and a collaborative thread bundle calculation method. The cooperative thread bundle calculation method comprises the following steps of: initializing a plurality of thread bundle units by a thread bundle resource allocation unit, so that the plurality of thread bundle units have the same initial vector register base address to form a group of cooperative thread bundles; the thread beam resource allocation unit notifies a thread beam scheduling and instruction transmitting unit to send a plurality of execution instructions to an execution unit; running a plurality of thread bundles of the plurality of thread bundle units by the execution unit according to the plurality of execution instructions; reading different scalar parameters from the respective scalar registers of the plurality of thread bundle units by the plurality of thread bundles; and accessing, by the plurality of thread bundles, the same register space of the vector register according to the same initial vector register base address. According to the artificial intelligence chip and the cooperative thread beam calculation method, the efficient cooperative thread beam calculation function of the cross-thread beam sharing vector register block can be achieved.

The invention provides a large language model operator compiling method oriented to a GPU platform, and belongs to the field of GPU operator compilation. A large language model operator is compiled through a Python interface to obtain a Python code; performing lexical analysis and grammatical analysis on the Python code to generate an abstract syntax tree; mapping the abstract syntax tree into an MMIR extension intermediate representation; tensor layout in the MLIR extension intermediate representation is uniformly converted into a binary linear mapping matrix, and optimization problems of global memory layout, shared memory layout and register block layout are analyzed based on the binary linear mapping matrix; automatically deducing a non-key tensor layout based on a program control flow diagram, generating a global optimal layout scheme, and obtaining an optimized MMIR extension intermediate representation; enabling the optimized MLIR extension intermediate representation to pass through a GGPU (Graphics Graphics Processing Unit) target code generator of the MLIR, and generating a PTX code adaptive to the NVIDIA GPU; the PTX code is loaded to a GPU, and an operator calculation result is output; the invention further provides a compiling system. The problem that the current large language model operator compiling efficiency is low is solved.

Embodiments of the present disclosure relate to applications, platforms, architecture, etc. for using a master test image that may be used for multiple different tests. For example, a testing system may include a register bank that may be loaded with test configurations corresponding to one or more tests. The test configurations may respectively correspond to sets of control packets included in the master test image that may be used or executed for corresponding tests. The test configurations may indicate execution orders of their respective sets of control packets in which the execution order of one or more of the control packets included in the sets of control packets may differ from a default execution order of such control packets as indicated in the master test image. Such a configuration may accordingly allow for the flexibility of performing many different tests using a single master test image.

The invention relates to the related technical field of reasoning calculation, in particular to an AI chip reasoning acceleration method and system and a medium, and the method comprises the steps of collecting adaptive scheduling parameters, setting a hierarchical pipeline response gradient under a reasoning acceleration framework, loading the optimized scheduling parameters to a configuration register group, and setting reasoning self-driven circulation. The technical problems that load feature perception is one-sided, scheduling parameters are statically solidified, dynamic adjustment along with reasoning tasks cannot be achieved, adaptability to feature graph structure changes and activation mode updating in the reasoning process is poor, and reasoning stability is insufficient are solved, and a tracing strategy based on a three-dimensional load feature vector and attention guidance is achieved. The method has the technical effects that the matching degree of scheduling parameters and task states is improved, the scheduling parameters and model compression granularity are dynamically adjusted in combination with real-time energy consumption constraints, a federated on-chip learning architecture enables the parameter iteration updating efficiency of a layer sensitivity analysis unit to be improved, the reasoning stability and the continuous operation reliability are enhanced, and the method adapts to end side complex dynamic scene requirements.

The application discloses a message encryption method and device, and relates to the technical field of data encryption. The message encryption method comprises the following steps of: in a previous clock cycle of processing adjacent first and second message words, a first hash module calculates a first e value and a first a value according to the first message word, a hash register value and a hash constant; a second hash module calculates a second e value and a second a value according to the second message word, the hash register value, the first e value, the first a value and the hash constant; and the second a value, the first a value, the second e value and the first e value are sequentially cached into a third hash register, a fourth hash register, a seventh hash register and an eighth hash register of a hash register group. Two hash modules are adopted to process two message words in the same clock cycle, the clock cycle used for calculating the message word digest value in the message encryption process is reduced, and therefore the user experience can be obviously improved.

The present invention provides a physical signal monitoring-based transmission link fault location method and system in the technical field of transmission link fault location. The method comprises: step S1, a layer 2 switching chip creating a register group, and an MCU creating a MAC address table; step S2, when a new network device is accessed, the MCU performing anomaly verification on a new MAC address by means of the MAC address table and then updating the MAC address table; step S3, performing fault location by means of a queue controller watchdog event, and updating an event register on the basis of a fault type; step S4, the MCU reading the fault type from the event register to match a fault handling measure from a fault management table; and step S5, when a fault handling measure is found, executing a transmission link recovery operation on the basis of the fault handling measure, and when no fault handling measure is found, reading switch data from the register group and uploading the switch data to a server for fault location, or executing fault location on the basis of a console command inputted from a Console interface. The present invention has the advantage that the convenience and timeliness of transmission link fault location are greatly improved.

The invention discloses a multichannel LED digital control system and method. The system comprises a main controller which generates an LED control instruction and sends a data frame through a differential bus; the plurality of LED driving modules are connected with the main controller through a differential bus, and each LED driving module comprises a differential receiver for receiving a differential signal of the differential bus; the UART communication module is connected with the differential receiver and converts the differential signal into serial data; the data processing state machine module is connected with the UART communication module and is used for analyzing the received data frame, and the data frame comprises a synchronous byte, a device address / command composite byte, a register address byte, a data byte and a CRC (Cyclic Redundancy Check) check byte; the digital logic counter is used for measuring the bit width of each data bit in the synchronous bytes and calibrating the sampling time sequence of the UART communication module based on a measurement result; the register group is used for storing LED control parameters; and the multi-path PWM generator is connected with the register block and generates multi-path PWM signals according to the LED control parameters so as to drive the LEDs. The power consumption is reduced.

The invention relates to the technical field of display, in particular to a display system capable of automatically brushing pictures, which comprises a TDDI chip integrating a main control function and a display driving function, and a storage module which is in communication connection with the TDDI chip and is used for storing background picture data and windowing picture data, a software application layer, a register block, a storage controller and an RGB time sequence generator are arranged in the TDDI chip, the software application layer configures automatic picture brushing enabling and related parameters to the register block, and the storage controller automatically reads picture data in a storage module according to configuration and achieves automatic switching between a background picture and a windowing picture. The RGB timing generator generates a display timing and drives the display panel. The system can also be selectively provided with an external MCU which is only used for image switching. Through the TDDI chip master control and display integrated design, an external master control does not need to participate in storage, reading and continuous transmission of image data, and the problems that a traditional display system is high in complexity, large in power consumption, high in master control requirement and large in resource occupation are solved.

The invention discloses an access address configuration method, a processor, a multiprocessor system, a medium and a product. The method comprises the following steps: configuring an independent register block for each thread bundle in the processor; wherein the register group is used for storing coordinate information of tensor data used in the execution process of a corresponding thread bundle program; reading the coordinate information of the tensor data to be processed from the register group, and sending the read coordinate information to a coprocessor to carry out access address calculation of the corresponding tensor data; according to the method, each thread bundle is provided with a group of entity registers for storing the coordinate information of the tensor data, multiple multiplexing of the coordinate information can be supported, and a large number of instructions for specifying coordinates are saved, so that the number of assembly instructions is reduced, the instruction overhead is effectively reduced, the execution time of a processor is saved, and the processing efficiency of the processor is improved.

The invention provides a processor system and an interrupt response method. In the processor system, an interrupt controller is used for setting the state of a control register, receiving an interrupt request signal which is sent from the outside and is used for requesting to execute an interrupt service program, setting the state of the control register to be a first state value when the interrupt request signal is received, and executing the interrupt service program after the interrupt server program is executed. Setting the state of the control register as a second state value; the multiplexer is used for reading the state value of the control register, when the read state value of the control register is a first state value, the multiplexer is communicated with the shadow register group and the arithmetic unit, and when the read state value of the control register is a second state value, the multiplexer is communicated with the main register group and the arithmetic unit; the main register group is used for storing data when the main program is operated; and the shadow register group is used for storing data when the interrupt server program runs. The delay in the interrupt response process can be reduced, and the system performance is improved.

The invention discloses an integrated hardware password hash calculation method based on SHA256 and SM3 algorithms. A message filling module, a message extension module and a compression iteration module are included. The message filling module is used for receiving externally input message data and performing unified message filling processing, and the message filling module is suitable for SHA256 and SM3 algorithms; the message expansion module comprises a Buffer with the depth of 16 and the bit width of 32 bits, and selects to execute an SHA256 or SM3 expansion function under the control of an sm3sel signal; and the compression iteration module completes 64 rounds or 68 rounds of iteration operation through an internal word register group and a compression function circuit, selects a corresponding algorithm according to sm3sel, and finally outputs a 256-bit hash result. According to the method, a uniform interface structure is adopted, HASH value calculation of an eMMC-RPMB module is supported, circuit resource reuse is achieved while compatibility is guaranteed, and chip area consumption is reduced.

Systems and techniques are provided for secure computing. For instances, a process can include generating a master private key; generating a set of first dummy values; dividing the master private key into a first set of shares, wherein a sum of shares of the first set of shares equals a value of the master private key; initiating portions of an active memory bank with a sequence of integer modular additions, wherein the integer modular additions comprise: masking the first set of shares of the master private key; and adding the masked first set of shares of the master private key to portions of the active memory bank using a sequence of adds that adds and mixes the masked first set of shares of the master private key with dummy values.

The invention belongs to the technical field of hardware design, and provides a sharing operation method, system and equipment of a network acceleration unit and a medium, and the method comprises the step of configuring a shared general register, an exclusive general register and an exclusive special register group for multiple hosts in a single network acceleration unit. Each host accesses the register through the low-speed bus based on the unique identifier. And the acceleration unit performs data interaction with each host through a high-speed bus according to the configuration of the register. When the host operation is processed, the management message and the data transmission message are distinguished by judging whether the basic communication resource register group is accessed or not, and high-order modification and resource mapping are carried out on a load data address in the data transmission message. When an external data packet is received, a target host is determined through analysis and address matching, and resource identifier mapping and data access are carried out. According to the invention, each host is ensured to have independent communication resources and data channels, the complexity and the cost of hardware implementation are reduced, and the resource utilization rate and the system expansibility are improved.

The invention relates to the technical field of computers, in particular to a rapid transposition method and device based on DMA (direct memory access) hardware, which uses a group of better configuration parameters, adopts source matrix parameters and target matrix parameters to describe three-dimensional data blocks and supports multi-batch matrix transposition. The block size matched with the transmission bit width of the preset communication protocol is adopted, data transmission is flexibly and efficiently completed, and the dynamic change problem in the data transposition process can be flexibly solved; part of source data is cached in the DMA hardware in advance for subsequent rapid extraction and use; matrix transposition of source data is achieved through the register set, data reading and writing are conducted in the row direction, the problem that a memory jumps when one column of the matrix is read is avoided under the condition that few hardware resources are used, the method is flexibly suitable for various transposition types, and transposition efficiency is improved.

A memory system includes a memory device, a control circuit, and a memory controller. The memory device includes a plurality of memory regions, a candidate register set storing a plurality of candidate parameter sets corresponding to temperature ranges, respectively, a mode register set storing a temperature range and a refresh management parameter set. The control circuit stores, based on the temperature range, a first candidate parameter set of the plurality of candidate parameter sets as the refresh management parameter set in the mode register set. The memory controller obtains the refresh management parameter set from the memory device, based on an updated temperature range, counts activation counts of the plurality of memory regions, and provides a refresh management command to the memory device based on the activation counts and the refresh management parameter set.

The invention provides a programmable logic device dynamic register configuration device and a control method thereof. The configuration device for the dynamic register of the programmable logic device comprises an external controller used for providing a configuration instruction and configuration data; the main controller is used for receiving and analyzing the configuration data, generating a gating signal and recombining a data packet; at least one multiplexing module, each multiplexing module being configured to include a register block; and one or more stages of sub-controllers are arranged between the main controller and the multiplexing module and are used for further analyzing the data packet and controlling the target register group to complete read-write operation. In addition, the invention further provides a control method corresponding to the configuration device. Some embodiments of the invention aim to solve the problems of high configuration switching time consumption and uncertain configuration switching time in the prior art on the premise of not increasing too much complexity and area overhead.

The application discloses an attention mechanism calculation method and device, a storage medium and a product. The method comprises the following steps: performing matrix multiplication operation on a query matrix block and a key matrix block to obtain a first product matrix block of a first precision type, performing exponential operation on the first product matrix block written in a second register group to obtain a numerator matrix block; writing the quantized numerator matrix block into a third register group according to a second precision type; writing the numerator matrix block into a fourth register group according to a third precision type and performing layout conversion through a shared memory; performing softmax operation on the layout-converted numerator matrix block and a value matrix block to obtain an attention result matrix block of the first precision type and write the attention result matrix block into a fifth register group; and writing the attention result matrix block into the shared memory according to the third precision type. The embodiment of the application can reduce the cumulative error of an intermediate operation result.

The invention relates to a processor comprising: an integer processing unit (ALU), a floating-point processing unit (FPU), a register bank (RGBK), an external data bus (DTB), a data path connecting the register bank to the floating-point processing unit, the integer processing unit, and the external data bus, and a processing chain configured to process at least partially simultaneous integer and floating-point instructions, the processing chain comprising common stages for processing an integer and a floating-point instruction, and at least one additional execution stage for executing a floating-point operation. Figure 1

A multicore processor may include a plurality of cores including at least a first core and a second core, a shared peripheral comprising a plurality of interrupt register banks including at least a first interrupt register bank dedicated to the first core and a second interrupt register bank dedicated to the second core, and a plurality of communications bridges, including at least a first bridge interfaced between the first core and the shared peripheral and at least a second bridge interfaced between the second core and the shared peripheral. The first core may be configured to program the first interrupt register bank via the first bridge to configure the shared peripheral for access by the first core. The second core may be configured to program the second interrupt register bank via the second bridge to configure the shared peripheral for access by the second core.

Control circuitry for memory includes a state machine including a number of state elements corresponding to a maximum number of available columns of a blast operation for memory; a set of registers including a corresponding register for each state element; and a column selection control circuit that combines outputs of the state machine and the set of registers to trigger an appropriate column during an appropriate clock cycle, available at a start of a corresponding clock cycle. The state machine receives a clock and various inputs associated with a start of memory operations and provides intermediate state element outputs and a final state element output as the outputs. Each register of the set of registers is available to store, from an address enable signal, a value indicating that a column in memory to which that register corresponds is to be accessed.

The invention relates to a processor comprising: an integer operation processing unit (ALU), a floating-point operation processing unit (FPU), a register bank (RGBK), an external data bus (DTB), a data path connecting the register bank to the floating-point operation processing unit, to the integer operation processing unit, and to the external data bus, and a processing chain configured to at least partially simultaneously process integer operation instructions and floating-point operation instructions, wherein the processing chain comprises common stages for processing an integer operation instruction and a floating-point operation instruction, and at least one additional execution stage for executing a floating-point operation.