Abstract: Disclosed embodiments relate to systems and methods for performing matrix row-wise and column-wise permute instructions. In one example, a processor includes fetch circuitry to fetch an instruction, decoding, using decode circuitry, the fetched instruction having fields to specify an opcode and locations of a source matrix and a destination matrix, the opcode indicating the processor is to perform a permutation by copying, into each of a plurality of equal-sized logical partitions of the destination matrix, a selected logical partition of a same size from the source matrix, the selection being indicated by a permute control, and execution circuitry to execute the decoded instruction as per the opcode.

Abstract: A system for processing data flow array instructions is described. The system includes a data flow array, which includes a plurality of processing elements; a decoder to receive a data flow array instruction and generate a set of microinstructions based on the data flow array instruction; a reservation station to receive and dispatch each microinstruction in the set of microinstructions, wherein the set of microinstructions includes a configuration microinstruction for configuring the data flow array for processing the data flow array instruction; a configuration watcher to receive the configuration microinstruction and to add a configuration identifier and a set of parameters of the configuration microinstruction to a configuration queue for the data flow array, wherein the data flow array is to configure the plurality of processing elements based on configuration information associated with the configuration identifier and the set of parameters.

Abstract: Disclosed embodiments relate to accelerating multiplication of sparse matrices. In one example, a processor is to fetch and decode an instruction having fields to specify locations of first, second, and third matrices, and an opcode indicating the processor is to multiply and accumulate matching non-zero (NZ) elements of the first and second matrices with corresponding elements of the third matrix, and executing the decoded instruction as per the opcode to generate NZ bitmasks for the first and second matrices, broadcast up to two NZ elements at a time from each row of the first matrix and each column of the second matrix to a processing engine (PE) grid, each PE to multiply and accumulate matching NZ elements of the first and second matrices with corresponding elements of the third matrix. Each PE further to store an NZ element for use in a subsequent multiplications.

Abstract: Disclosed embodiments relate to systems and methods for performing nibble-sized operations on matrix elements. In one example, a processor includes fetch circuitry to fetch an instruction, decode circuitry to decode the fetched instruction the fetched instruction having fields to specify an opcode and locations of first source, second source, and destination matrices, the opcode to indicate the processor is to, for each pair of corresponding elements of the first and second source matrices, logically partition each element into nibble-sized partitions, perform an operation indicated by the instruction on each partition, and store execution results to a corresponding nibble-sized partition of a corresponding element of the destination matrix. The exemplary processor includes execution circuitry to execute the decoded instruction as per the opcode.

Abstract: Systems, methods, and apparatuses relating to performing fast Fourier transform (FFT) configuration and computation operations are described. In one embodiment, a processor includes a matrix operations accelerator circuit that includes a two-dimensional grid of processing element circuits; a first plurality of registers that represents a first two-dimensional matrix coupled to the matrix operations accelerator circuit; a second plurality of registers that represents a second two-dimensional matrix coupled to the matrix operations accelerator circuit; a decoder, of a core coupled to the matrix operations accelerator circuit, to decode a single instruction into a decoded single instruction; and an execution circuit of the core to execute the decoded single instruction to cause the two-dimensional grid of processing element circuits to operate on a first packed data input value and a first complex twiddle factor value to produce a first result and a second result.

Abstract: Systems, methods, and apparatuses relating to performing stencil configuration and computation operations are described.

Abstract: An apparatus and method for efficiently performing a multiply add or multiply accumulate operation.

Abstract: An apparatus and method for efficient matrix alignment in a systolic array.

Abstract: Disclosed embodiments relate to instructions for fast element unpacking. In one example, a processor includes fetch circuitry to fetch an instruction whose format includes fields to specify an opcode and locations of an Array-of-Structures (AOS) source matrix and one or more Structure of Arrays (SOA) destination matrices, wherein: the specified opcode calls for unpacking elements of the specified AOS source matrix into the specified Structure of Arrays (SOA) destination matrices, the AOS source matrix is to contain N structures each containing K elements of different types, with same-typed elements in consecutive structures separated by a stride, the SOA destination matrices together contain K segregated groups, each containing N same-typed elements, decode circuitry to decode the fetched instruction, and execution circuitry, responsive to the decoded instruction, to unpack each element of the specified AOS matrix into one of the K element types of the one or more SOA matrices.

Abstract: Disclosed embodiments relate to matrix compress/decompress instructions. In one example, a processor includes fetch circuitry to fetch a compress instruction having a format with fields to specify an opcode and locations of decompressed source and compressed destination matrices, decode circuitry to decode the fetched compress instructions, and execution circuitry, responsive to the decoded compress instruction, to: generate a compressed result according to a compress algorithm by compressing the specified decompressed source matrix by either packing non-zero-valued elements together and storing the matrix position of each non-zero-valued element in a header, or using fewer bits to represent one or more elements and using the header to identify matrix elements being represented by fewer bits; and store the compressed result to the specified compressed destination matrix.

Abstract: According to some embodiments, first and second processing elements may be provided on a die, and there may be a plurality of potential communication links between the first and second processing elements. Moreover, control logic may be provided on the die to dynamically activate at least some of the potential communication links (e.g., based on a current bandwidth appropriate between the first and second processing elements).

Abstract: Methods and apparatuses for error correction. A N-bit block data to be stored in a memory device is received. The memory device does not perform any error correction code (ECC) algorithm nor provide designated error correction code storage for the N-bit block of data. Data compression is applied to the N-bit data to compress the block of data to generate a M-bit compressed block of data. A K-bit ECC is computed for the M-bit compressed data, wherein M+K is less than or equal to N. The M-bit compressed data and the K-bit ECC are stored together in the memory device.

Abstract: Methods and apparatuses for error correction. A N-bit block data to be stored in a memory device is received. The memory device does not perform any error correction code (ECC) algorithm nor provide designated error correction code storage for the N-bit block of data. Data compression is applied to the N-bit data to compress the block of data to generate a M-bit compressed block of data. A K-bit ECC is computed for the M-bit compressed data, wherein M+K is less than or equal to N. The M-bit compressed data and the K-bit ECC are stored together in the memory device.

Abstract: According to some embodiments, first and second processing elements may be provided on a die, and there may be a plurality of potential communication links between the first and second processing elements. Moreover, control logic may be provided on the die to dynamically activate at least some of the potential communication links (e.g., based on a current bandwidth appropriate between the first and second processing elements).