Log in
Sign up
Golden has been acquired by ComplyAdvantage.Read about it here ⟶
‌

US Patent 11842423 Dot product operations on sparse matrix elements

Patent 11842423 was granted and assigned to Intel on December, 2023 by the United States Patent and Trademark Office.

OverviewStructured DataIssuesContributorsActivity

All edits

Edits on 13 Dec, 2023
"Created via: Patent importer"
Golden AI profile picture
Golden AI
created this topic on 13 Dec, 2023
Edits made to:
Infobox (+77 properties)
Article (+397 characters)
‌

 US Patent 11842423 Dot product operations on sparse matrix elements

Article

Patent abstract

Embodiments described herein include software, firmware, and hardware logic that provides techniques to perform arithmetic on sparse data via a systolic processing unit. One embodiment provides for data aware sparsity via compressed bitstreams. One embodiment provides for block sparse dot product instructions. One embodiment provides for a depth-wise adapter for a systolic array.

Infobox
Is a
Patent
Patent
0
Patent Applicant
Intel
Intel
0
Current Assignee
Intel
Intel
0
Patent Jurisdiction
United States Patent and Trademark Office
United States Patent and Trademark Office
0
Patent Number
118424230
Patent Inventor Names
Abhishek Appu0
Ashutosh Garg0
Fangwen Fu0
Jiasheng Chen0
Joydeep Ray0
Mike Macpherson0
Subramaniam Maiyuran0
Varghese George0
Vasanth Ranganathan0
Date of Patent
December 12, 2023
0
Patent Application Number
171229050
Date Filed
December 15, 2020
0
Patent Citations
‌
US Patent 11113784 Sparse optimizations for a matrix accelerator architecture
0
‌
US Patent 11080046 Instructions and logic to perform floating point and integer operations for machine learning
0
‌
US Patent 11169799 Instructions and logic to perform floating-point and integer operations for machine learning
0
‌
US Patent 7102646 Demand-based memory system for graphics applications
0
‌
US Patent 7127482 Performance optimized approach for efficient downsampling operations
0
‌
US Patent 7197605 Allocating cache lines
0
‌
US Patent 7327289 Data-modifying run length encoder to avoid data expansion
0
‌
US Patent 7346741 Memory latency of processors with configurable stride based pre-fetching technique
0
‌
US Patent 7373369 Advanced execution of extended floating-point add operations in a narrow dataflow
0
‌
US Patent 7483031 Method for synchronizing graphics processing units
0
‌
US Patent 7616206 Efficient multi-chip GPU
0
‌
US Patent 7620793 Mapping memory partitions to virtual memory pages
0
‌
US Patent 7873812 Method and system for efficient matrix multiplication in a SIMD processor architecture
0
‌
US Patent 7913041 Cache reconfiguration based on analyzing one or more characteristics of run-time performance data or software hint
0
‌
US Patent 8253750 Digital media processor
0
‌
US Patent 8340280 Using a single instruction multiple data (SIMD) instruction to speed up galois counter mode (GCM) computations
0
‌
US Patent 8429351 Techniques for determining an amount of data to prefetch
0
‌
US Patent 8488055 Flash synchronization using image sensor interface timing signal
0
‌
US Patent 8669990 Sharing resources between a CPU and GPU
0
‌
US Patent 8847965 Methods, systems, and computer readable media for fast geometric sound propagation using visibility computations
0
‌
US Patent 8990505 Cache memory bank selection
0
‌
US Patent 9304835 Optimized system for analytics (graphs and sparse matrices) operations
0
‌
US Patent 9317251 Efficient correction of normalizer shift amount errors in fused multiply add operations
0
‌
US Patent 9461667 Rounding injection scheme for floating-point to integer conversion
0
‌
US Patent 9501392 Management of a non-volatile memory module
0
‌
US Patent 9558156 Sparse matrix multiplication using a single field programmable gate array module
0
‌
US Patent 9727337 Fine-grained instruction enablement at sub-function granularity based on an indicated subrange of registers
0
‌
US Patent 9804666 Warp clustering
0
‌
US Patent 9811468 Set associative cache memory with heterogeneous replacement policy
0
‌
US Patent 9960917 Matrix multiply accumulate instruction
0
‌
US Patent 10102015 Just in time GPU executed program cross compilation
0
‌
US Patent 10353706 Instructions and logic to perform floating-point and integer operations for machine learning
0
‌
US Patent 10409614 Instructions having support for floating point and integer data types in the same register
0
‌
US Patent 10409887 Generalized dot product for computer vision applications
0
‌
US Patent 10474458 Instructions and logic to perform floating-point and integer operations for machine learning
0
‌
US Patent 10528864 Sparse convolutional neural network accelerator
0
‌
US Patent 10572409 Sparse matrix processing circuitry
0
‌
US Patent 10678508 Accelerated quantized multiply-and-add operations
0
‌
US Patent 10755201 Systems and methods for data collection and analysis at the edge
0
‌
US Patent 10762137 Page table search engine
0
‌
US Patent 10762164 Vector and matrix computing device
0
‌
US Patent 10860316 Apparatus and methods for generating dot product
0
‌
US Patent 10860922 Sparse convolutional neural network accelerator
0
‌
US Patent 10891538 Sparse convolutional neural network accelerator
0
‌
US Patent 10896045 Architecture for dense operations in machine learning inference engine
0
‌
US Patent 11250108 Matrix processing method and apparatus, and logic circuit
0
‌
US Patent 11361496 Graphics processors and graphics processing units having dot product accumulate instruction for hybrid floating point format
0
‌
US Patent 11360767 Instructions and logic to perform floating point and integer operations for machine learning
0
‌
US Patent 11409537 Mixed inference using low and high precision
0
‌
US Patent 11461107 Compute unit having independent data paths
0
‌
US Patent 11620256 Systems and methods for improving cache efficiency and utilization
0
‌
US Patent 6856320 Demand-based memory system for graphics applications
0
‌
US Patent 6947049 Method and system for synchronizing updates of vertex data with a graphics processor that is fetching vertex data
0
Patent Primary Examiner
‌
Shawn Doman
0
CPC Code
‌
G06F 15/8046
0
‌
G06F 9/5027
0
‌
G06F 7/5443
0
‌
G06N 3/08
0
‌
G06N 3/0481
0
‌
G06F 17/16
0

Find more entities like US Patent 11842423 Dot product operations on sparse matrix elements

Use the Golden Query Tool to find similar entities by any field in the Knowledge Graph, including industry, location, and more.
Open Query Tool
Access by API
Golden Query Tool
Golden logo
By using this site, you agree to our Terms of Service.