Abstract: Embodiments detailed herein relate to arithmetic operations of float-point values. An exemplary processor includes decoding circuitry to decode an instruction, where the instruction specifies locations of a plurality of operands, values of which being in a floating-point format. The exemplary processor further includes execution circuitry to execute the decoded instruction, where the execution includes to: convert the values for each operand, each value being converted into a plurality of lower precision values, where an exponent is to be stored for each operand; perform arithmetic operations among lower precision values converted from values for the plurality of the operands; and generate a floating-point value by converting a resulting value from the arithmetic operations into the floating-point format and store the floating-point value.

Abstract: Methods of encoding and decoding are described which use a variable number of instruction words to encode instructions from an instruction set, such that different instructions within the instruction set may be encoded using different numbers of instruction words. To encode an instruction, the bits within the instruction are re-ordered and formed into instruction words based upon their variance as determined using empirical or simulation data. The bits in the instruction words are compared to corresponding predicted values and some or all of the instruction words that match the predicted values are omitted from the encoded instruction.

Abstract: In one embodiment, a method includes accessing a loaded but paused source process executable and disassembling the source process executable to identify a system call to be instrumented and an adjacent relocatable instruction. Instrumenting the system call includes building a trampoline for the system call that includes a check flag instruction at or near an entry point to the trampoline and two areas of the trampoline that are selectively executed according to results of the check flag instruction. Building a first area of the trampoline includes providing instructions to execute a relocated copy of the adjacent relocatable instruction and return flow to an address immediately following the adjacent relocatable instruction. Building a second area of the trampoline includes providing instructions to invoke at least one handler associated with executing a relocated copy of the system call and return flow to an address immediately following the system call.

Abstract: A method and apparatus for controlling pre-fetching in a processor. A processor includes an execution pipeline and an instruction pre-fetch unit. The execution pipeline is configured to execute instructions. The instruction pre-fetch unit is coupled to the execution pipeline. The instruction pre-fetch unit includes instruction storage to store pre-fetched instructions, and pre-fetch control logic. The pre-fetch control logic is configured to fetch instructions from memory and store the fetched instructions in the instruction storage. The pre-fetch control logic is also configured to provide instructions stored in the instruction storage to the execution pipeline for execution. The pre-fetch control logic is further configured set a maximum number of instruction words to be pre-fetched for execution subsequent to execution of an instruction currently being executed in the execution pipeline.

Abstract: A software-based instruction scoreboard indicates dependencies between closely-issued instructions issued to an arithmetic logic unit (ALU) pipeline. The software-based instruction scoreboard inserts one or more control words into the command stream between the dependent instructions, which is then executed by the ALU pipeline. The control words identify the instruction(s) upon which the dependent instructions depend (parent instructions) so that the GPU hardware can ensure that the ALU pipeline does not stall while the dependent instruction waits for results from the parent instruction.

Abstract: Obsoleting values stored in registers in a processor based on processing obsolescent register-encoded instructions is disclosed. The processor is configured to support execution of read and/or write instructions that include obsolescence encoding indicating that one or more of its source and/or target register operands are to be obsoleted by the processor. A register encoded as obsolescent means the data value stored in such register will not be used by subsequent instructions in an instruction stream, and thus does not need to be retained. Thus, such register can be set as being in an obsolescent state so that the data value stored in such register can be ignored to improve performance. As one example, data values for registers having an obsolescent state can be ignored and thus not stored in a saved context for a process being switched out, thus conserving memory and improving processing time for a process switch.

Abstract: An arithmetic processing device, includes a memory; and a processor coupled to the memory and the processor configured to: execute arithmetic processing which executes a plurality of instructions issued out of order, execute control processing which commits the plurality of instructions for which execution has been completed in order, identify, for each instruction included in the plurality of instructions, a count value which indicates a number of cycles from when execution of the instruction has been completed, and identify, among a plurality of uncommitted instructions, the instruction with the count value which matches the number of cycles in which an error is detected in the arithmetic processing as a specific instruction to be retried.

Abstract: A neural processing device is provided. The neural processing device comprises: a processing unit configured to receive an input activation and a weight and perform a two-dimensional matrix calculation with the input activation and the weight to generate an output activation, a first memory, and a load-store unit (LSU) configured to perform memory access operations between the first memory and a second memory. The memory access operations include a main memory access operation for a current processing operation that is performed by the processing unit, and a standby memory access operation for a standby processing operation that is performed by the processing unit after the current processing operation. A level of the first memory is equal to a level of the processing unit, and a level of the second memory is different from the level of the first memory.

Abstract: A predictor includes a memory having a plurality of entries. Each entry includes a prediction of a hash of a next fetch address produced by a fetch block J of a series of successive fetch blocks in program execution order and a branch direction produced by the fetch block J. An input selects an entry for provision on the output. The output is fed back to the input such that the output provides the prediction of the hash of the next fetch address and the branch direction produced by each fetch block over a series of successive clock cycles. The hash of the next fetch address is insufficient for use by an instruction fetch unit to fetch from an instruction cache a fetch block J+1, whereas the next fetch address itself is sufficient for use by the instruction fetch unit to fetch from the instruction cache the fetch block J+1.

Abstract: Representative apparatus, method, and system embodiments are disclosed for a self-scheduling processor which also provides additional functionality. Representative embodiments include a self-scheduling processor, comprising: a processor core adapted to execute a received instruction; and a core control circuit adapted to automatically schedule an instruction for execution by the processor core in response to a received work descriptor data packet. In another embodiment, the core control circuit is also adapted to schedule a fiber create instruction for execution by the processor core, to reserve a predetermined amount of memory space in a thread control memory to store return arguments, and to generate one or more work descriptor data packets to another processor or hybrid threading fabric circuit for execution of a corresponding plurality of execution threads. Event processing, data path management, system calls, memory requests, and other new instructions are also disclosed.

Abstract: Representative apparatus, method, and system embodiments are disclosed for a self-scheduling processor which also provides additional functionality. Representative embodiments include a self-scheduling processor, comprising: a processor core adapted to execute a received instruction; and a core control circuit adapted to automatically schedule an instruction for execution by the processor core in response to a received work descriptor data packet. In another embodiment, the core control circuit is also adapted to schedule a fiber create instruction for execution by the processor core, to reserve a predetermined amount of memory space in a thread control memory to store return arguments, and to generate one or more work descriptor data packets to another processor or hybrid threading fabric circuit for execution of a corresponding plurality of execution threads. Event processing, data path management, system calls, memory requests, and other new instructions are also disclosed.

Abstract: Representative apparatus, method, and system embodiments are disclosed for a self-scheduling processor which also provides additional functionality. Representative embodiments include a self-scheduling processor, comprising: a processor core adapted to execute a received instruction; and a core control circuit adapted to automatically schedule an instruction for execution by the processor core in response to a received work descriptor data packet. In another embodiment, the core control circuit is also adapted to schedule a fiber create instruction for execution by the processor core, to reserve a predetermined amount of memory space in a thread control memory to store return arguments, and to generate one or more work descriptor data packets to another processor or hybrid threading fabric circuit for execution of a corresponding plurality of execution threads. Event processing, data path management, system calls, memory requests, and other new instructions are also disclosed.

Abstract: An apparatus and method are provided for processing instructions. The apparatus has execution circuitry for executing instructions, where each instruction requires an associated operation to be performed using one or more source operand values in order to produce a result value. Issue circuitry is used to maintain a record of pending instructions awaiting execution by the execution circuitry, and prediction circuitry is used to produce a predicted source operand value for a chosen pending instruction. Optimisation circuitry is then arranged to detect an optimisation condition for the chosen pending instruction when the predicted source operand value is such that, having regard to the associated operation for the chosen pending instruction, the result value is known without performing the associated operation.

Abstract: The apparatus and method for calculating and retaining a bound on error during floating-point operations inserts an additional bounding field into the standard floating-point format that records the retained significant bits of the calculation with notification upon insufficient retention. The bounding field, accounting for both rounding and cancellation errors, includes the lost bits D Field and the accumulated rounding error R Field. The D Field states the number of bits in the floating-point representation that are no longer meaningful. The bounds on the represented real value are determined by the truncated floating-point value and the addition of the error determined by the number of lost bits. The true, real value is absolutely contained by these bounds. The allowable loss (optionally programmable) of significant digits provides a fail-safe, real-time notification of loss of significant digits. This allows representation of real numbers accurate to the last digit.

Abstract: Disclosed embodiments relate to computing dot products of nibbles in tile operands. In one example, a processor includes decode circuitry to decode a tile dot product instruction having fields for an opcode, a destination identifier to identify a M by N destination matrix, a first source identifier to identify a M by K first source matrix, and a second source identifier to identify a K by N second source matrix, each of the matrices containing doubleword elements, and execution circuitry to execute the decoded instruction to perform a flow K times for each element (M,N) of the identified destination matrix to generate eight products by multiplying each nibble of a doubleword element (M,K) of the identified first source matrix by a corresponding nibble of a doubleword element (K,N) of the identified second source matrix, and to accumulate and saturate the eight products with previous contents of the doubleword element (M,N).

Abstract: Multi-level hierarchical routing matrices for pattern-recognition processors are provided. One such routing matrix may include one or more programmable and/or non-programmable connections in and between levels of the matrix. The connections may couple routing lines to feature cells, groups, rows, blocks, or any other arrangement of components of the pattern-recognition processor.

Abstract: An apparatus and method are described for a multithreaded-aware performance monitor of a processor. For example, one embodiment of a processor comprises: one or more simultaneous multithreading cores to simultaneously execute multiple instruction threads; a plurality of performance monitor counters, each performance monitor counter to count baseline events during processing of the multiple instruction threads; and a performance monitor circuit to determine whether multiple threads are concurrently generating the same baseline event and, if so, then the performance monitor circuit to distribute the count of the baseline event for only one of the multiple threads in each processor cycle for which the multiple threads are active and the baseline event applies to.

Abstract: The present disclosure is directed to systems and methods for a Processing-In-Memory Device that is configured to perform dot product calculations. A sequence control may be used to store data in a memory array according to an allocation pattern. The cells of the memory array may correspond to array elements of the data. The sequence control may apply another array of data to groups of elements within the memory array using the allocation pattern to perform dot product calculations. The dot product calculations may be used, for example, to implement a layer in a convolutional neural network.

Abstract: Aspects of the present disclosure relate an apparatus comprising fetch circuitry and instruction storage circuitry. The fetch circuitry is to fetch instructions for execution by execution circuitry. The instruction storage circuitry is to store temporary copies of fetched instructions. The fetch circuitry is configured to preferentially fetch instructions from the instruction storage circuitry. The instruction storage circuitry is configured to, responsive to a storage condition being met, begin storing copies of consecutive fetched instructions, the storage condition indicating a utility of a current fetched instruction; and to, responsive to determining that a number of said stored consecutive instructions has reached a storage threshold, cease storing copies of subsequent fetched instructions.

Abstract: An exception summary is provided for an invalid value detected during instruction execution. An indication that a value determined to be invalid was included in input data to a computation of one or more computations or in output data resulting from the one or more computations is obtained. The value is determined to be invalid due to one exception of a plurality of exceptions. Based on obtaining the indication that the value is determined to be invalid, a summary indicator is set. The summary indicator represents the plurality of exceptions collectively.