Accelerator for sparse-dense matrix multiplication

S Narayanamoorthy, NR Satish, A Suprun… - US Patent …, 2020‏ - Google Patents
Disclosed embodiments relate to an accelerator for sparse dense matrix instructions. In one
example, a processor to execute a sparse-dense matrix multiplication instruction, includes …

Systems and methods for exchange of data in distributed training of machine learning algorithms

A Matveev, N Shavit - US Patent 11,715,287, 2023‏ - Google Patents
Systems and methods may make exchanging data in a neural network (NN) during training
more eficient. Exchanging weights among a number of processors training a NN across …

Systems and methods for improved neural network execution

A Matveev, N Shavit - US Patent 11,449,363, 2022‏ - Google Patents
A method and system for computing one or more outputs of a neural network having a
plurality of layers is provided. The method and system can include determining a plurality of …

Methods and systems for improved transforms in convolutional neural networks

A Matveev, N Shavit - US Patent 10,902,318, 2021‏ - Google Patents
US10902318B2 - Methods and systems for improved transforms in convolutional neural networks
- Google Patents US10902318B2 - Methods and systems for improved transforms in convolutional …

Sparse matrix multiplication using a single field programmable gate array module

C Bekas, A Curioni, H Giefers, C Hagleitner… - US Patent …, 2020‏ - Google Patents
According to some embodiments, a computer-implemented method for performing sparse
matrix dense matrix (SPMM) multiplication on a single field programmable gate array …

System and method for executing convolution in a neural network

J Kopinsky - US Patent 11,544,559, 2023‏ - Google Patents
(57) ABSTRACT A system and method of executing a convolution layer of a neural network
may include:(a) selecting an output spatial position (OSP) of an output matrix data element …

Dot product operations on sparse matrix elements

A Appu, S Maiyuran, M MacPherson, F Fu… - US Patent …, 2023‏ - Google Patents
2020-12-16 Assigned to INTEL CORPORATION reassignment INTEL CORPORATION
ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS). Assignors …

Hardware accelerator for compressed LSTM

S Han, XIE Dongliang, K Junlong, Y Li - US Patent 10,691,996, 2020‏ - Google Patents
Hardware accelerator for compressed Long Short Term Memory (LSTM) is disclosed. The
accelerator comprise a sparse matrix-vector multiplication module for performing …

Graphics processor operation scheduling for deterministic latency

J Ray, S Panneer, S Tangri, B Ashbaugh… - US Patent …, 2024‏ - Google Patents
Embodiments described herein include software, firmware, and hardware that provides
techniques to enable deterministic scheduling across multiple general-purpose graphics …

Instruction based control of memory attributes

J Ray, A Koker, V George, M MacPherson… - US Patent …, 2024‏ - Google Patents
Embodiments described herein provide techniques to facilitate instruction-based control of
memory attributes. One embodiment provides a graphics processor comprising a processing …