Patents by Inventor Ashutosh Garg
Ashutosh Garg has filed for patents to protect the following inventions. This listing includes patent applications that are pending as well as patents that have already been granted by the United States Patent and Trademark Office (USPTO).
-
Patent number: 11954063Abstract: Described herein is a graphics processing unit (GPU) configured to receive an instruction having multiple operands, where the instruction is a single instruction multiple data (SIMD) instruction configured to use a bfloat16 (BF16) number format and the BF16 number format is a sixteen-bit floating point format having an eight-bit exponent. The GPU can process the instruction using the multiple operands, where to process the instruction includes to perform a multiply operation, perform an addition to a result of the multiply operation, and apply a rectified linear unit function to a result of the addition.Type: GrantFiled: February 17, 2023Date of Patent: April 9, 2024Assignee: Intel CorporationInventors: Subramaniam Maiyuran, Shubra Marwaha, Ashutosh Garg, Supratim Pal, Jorge Parra, Chandra Gurram, Varghese George, Darin Starkey, Guei-Yuan Lueh
-
Publication number: 20240053985Abstract: Embodiments described herein provide an apparatus comprising a plurality of processing resources including a first processing resource and a second processing resource, a shared local memory communicatively coupled to the first processing resource and the second processing resource, and a processor to receive an instruction to initiate a matrix multiplication operation, write a first set of matrix data into a first set of registers, and share the first set of matrix data between the first processing resource and the second processing resource for use in the matrix multiplication operation. Other embodiments may be described and claimed.Type: ApplicationFiled: October 11, 2023Publication date: February 15, 2024Applicant: Intel CorporationInventors: SUBRAMANIAM MAIYURAN, VARGHESE GEORGE, JOYDEEP RAY, ASHUTOSH GARG, JORGE PARRA, SHUBH SHAH, SHUBRA MARWAHA
-
Patent number: 11900502Abstract: Examples described herein relate to a software and hardware optimization that manages scenarios where a write operation to a register is less than an entirety of the register. A compiler detects instructions that make partial writes to the same register, groups such instructions, and provides hints to hardware of the partial write. The execution unit combines the output data for grouped instructions and updates the destination register as single write instead of multiple separate partial writes.Type: GrantFiled: May 2, 2022Date of Patent: February 13, 2024Assignee: Intel CorporationInventors: Chandra S. Gurram, Gang Y. Chen, Subramaniam Maiyuran, Supratim Pal, Ashutosh Garg, Jorge E. Parra, Darin M. Starkey, Guei-Yuan Lueh, Wei-Yu Chen
-
Patent number: 11842423Abstract: 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.Type: GrantFiled: December 15, 2020Date of Patent: December 12, 2023Assignee: Intel CorporationInventors: Abhishek Appu, Subramaniam Maiyuran, Mike Macpherson, Fangwen Fu, Jiasheng Chen, Varghese George, Vasanth Ranganathan, Ashutosh Garg, Joydeep Ray
-
Publication number: 20230351543Abstract: Embodiments described herein include, software, firmware, and hardware logic that provides techniques to perform arithmetic on sparse data via a systolic processing unit. Embodiment described herein provided techniques to detect zero value elements within a vector or a set of packed data elements output by a processing resource and generate metadata to indicate a location of the zero value elements within the plurality of data elements.Type: ApplicationFiled: May 2, 2023Publication date: November 2, 2023Applicant: Intel CorporationInventors: Joydeep Ray, Scott Janus, Varghese George, Subramaniam Maiyuran, Altug Koker, Abhishek Appu, Prasoonkumar Surti, Vasanth Ranganathan, Valentin Andrei, Ashutosh Garg, Yoav Harel, Arthur Hunter, JR., SungYe Kim, Mike Macpherson, Elmoustapha Ould-Ahmed-Vall, William Sadler, Lakshminarayanan Striramassarma, Vikranth Vemulapalli
-
Publication number: 20230297373Abstract: Embodiments described herein provided for an instruction and associated logic to enable GPGPU program code to access special purpose hardware logic to accelerate dot product operations. One embodiment provides for a graphics processing unit comprising a fetch unit to fetch a single instruction for execution, a decode unit to decode the single instruction into a decoded instruction, wherein the decoded instruction is to cause the graphics processing unit to perform a set of parallel dot product operations on elements of input matrices, and a systolic dot product unit to execute the decoded instruction across one or more parallel processor lanes using multiple systolic layers associated with multiple pipeline stages. The multiple pipeline stages include one or more sets of interconnected multipliers and adders to compute multiple concurrent dot products.Type: ApplicationFiled: April 26, 2023Publication date: September 21, 2023Applicant: Intel CorporationInventors: SUBRAMANIAM MAIYURAN, GUEI-YUAN LUEH, SUPRATIM PAL, ASHUTOSH GARG, CHANDRA S. GURRAM, JORGE E. PARRA, JUNJIE GU, KONRAD TRIFUNOVIC, HONG BIN LIAO, MIKE B. MACPHERSON, SHUBH B. SHAH, SHUBRA MARWAHA, STEPHEN JUNKINS, TIMOTHY R. BAUER, VARGHESE GEORGE, WEIYU CHEN
-
Publication number: 20230281272Abstract: Described herein is a graphics processor including a plurality of processing clusters coupled with a host interface, each processing cluster comprising a plurality of multiprocessors, the plurality of multiprocessors interconnected via a data interconnect, and each multiprocessor comprising sparse matrix multiply acceleration hardware including a systolic processing array with feedback inputs.Type: ApplicationFiled: April 17, 2023Publication date: September 7, 2023Applicant: Intel CorporationInventors: SUBRAMANIAM MAIYURAN, JORGE PARRA, SUPRATIM PAL, ASHUTOSH GARG, SHUBRA MARWAHA, CHANDRA GURRAM, DARIN STARKEY, DURGESH BORKAR, VARGHESE GEORGE
-
Patent number: 11727201Abstract: A system and method for transferring annotations associated with a media file. An annotation associated with a media file is indexed to a first instance of that media file. By comparing features of the two instances, a mapping is created between the first instance of the media file and a second instance of the media file. The annotation can be indexed to the second instance using the mapping between the first and second instances. The annotation can be processed (displayed, stored, or modified) based on the index to the second instance.Type: GrantFiled: August 22, 2022Date of Patent: August 15, 2023Assignee: Google LLCInventors: Mayur Datar, Ashutosh Garg, Vibhu Mittal
-
Patent number: 11709793Abstract: Described herein is a graphics processing unit (GPU) comprising a first processing cluster to perform parallel processing operations, the parallel processing operations including a ray tracing operation and a matrix multiply operation; and a second processing cluster coupled to the first processing cluster, wherein the first processing cluster includes a floating-point unit to perform floating point operations, the floating-point unit is configured to process an instruction using a bfloat16 (BF16) format with a multiplier to multiply second and third source operands while an accumulator adds a first source operand with output from the multiplier.Type: GrantFiled: May 27, 2022Date of Patent: July 25, 2023Assignee: Intel CorporationInventors: Subramaniam Maiyuran, Shubra Marwaha, Ashutosh Garg, Supratim Pal, Jorge Parra, Chandra Gurram, Varghese George, Darin Starkey, Guei-Yuan Lueh
-
Publication number: 20230195685Abstract: Described herein is a graphics processing unit (GPU) configured to receive an instruction having multiple operands, where the instruction is a single instruction multiple data (SIMD) instruction configured to use a bfloat16 (BF16) number format and the BF16 number format is a sixteen-bit floating point format having an eight-bit exponent. The GPU can process the instruction using the multiple operands, where to process the instruction includes to perform a multiply operation, perform an addition to a result of the multiply operation, and apply a rectified linear unit function to a result of the addition.Type: ApplicationFiled: February 17, 2023Publication date: June 22, 2023Applicant: Intel CorporationInventors: Subramaniam Maiyuran, Shubra Marwaha, Ashutosh Garg, Supratim Pal, Jorge Parra, Chandra Gurram, Varghese George, Darin Starkey, Guei-Yuan Lueh
-
Patent number: 11676239Abstract: Embodiments described herein include, software, firmware, and hardware logic that provides techniques to perform arithmetic on sparse data via a systolic processing unit. Embodiment described herein provided techniques to skip computational operations for zero filled matrices and sub-matrices. Embodiments additionally provide techniques to maintain data compression through to a processing unit. Embodiments additionally provide an architecture for a sparse aware logic unit.Type: GrantFiled: June 3, 2021Date of Patent: June 13, 2023Assignee: Intel CorporationInventors: Joydeep Ray, Scott Janus, Varghese George, Subramaniam Maiyuran, Altug Koker, Abhishek Appu, Prasoonkumar Surti, Vasanth Ranganathan, Andrei Valentin, Ashutosh Garg, Yoav Harel, Arthur Hunter, Jr., SungYe Kim, Mike Macpherson, Elmoustapha Ould-Ahmed-Vall, William Sadler, Lakshminarayanan Striramassarma, Vikranth Vemulapalli
-
Patent number: 11669329Abstract: Embodiments described herein provide for an instruction and associated logic to enable a vector multiply add instructions with automatic zero skipping for sparse input. One embodiment provides for a general-purpose graphics processor comprising logic to perform operations comprising fetching a hardware macro instruction having a predicate mask, a repeat count, and a set of initial operands, where the initial operands include a destination operand and multiple source operands. The hardware macro instruction is configured to perform one or more multiply/add operations on input data associated with a set of matrices.Type: GrantFiled: April 18, 2022Date of Patent: June 6, 2023Assignee: Intel CorporationInventors: Supratim Pal, Sasikanth Avancha, Ishwar Bhati, Wei-Yu Chen, Dipankar Das, Ashutosh Garg, Chandra S. Gurram, Junjie Gu, Guei-Yuan Lueh, Subramaniam Maiyuran, Jorge E. Parra, Sudarshan Srinivasan, Varghese George
-
Patent number: 11640297Abstract: Embodiments described herein provided for an instruction and associated logic to enable GPGPU program code to access special purpose hardware logic to accelerate dot product operations. One embodiment provides for a graphics processing unit comprising a fetch unit to fetch an instruction for execution and a decode unit to decode the instruction into a decoded instruction. The decoded instruction is a matrix instruction to cause the graphics processing unit to perform a parallel dot product operation. The GPGPU also includes systolic dot product circuitry to execute the decoded instruction across one or more SIMD lanes using multiple systolic layers, wherein to execute the decoded instruction, a dot product computed at a first systolic layer is to be output to a second systolic layer, wherein each systolic layer includes one or more sets of interconnected multipliers and adders, each set of multipliers and adders to generate a dot product.Type: GrantFiled: June 15, 2021Date of Patent: May 2, 2023Assignee: Intel CorporationInventors: Subramaniam Maiyuran, Guei-Yuan Lueh, Supratim Pal, Ashutosh Garg, Chandra S. Gurram, Jorge E. Parra, Junjie Gu, Konrad Trifunovic, Hong Bin Liao, Mike B. MacPherson, Shubh B. Shah, Shubra Marwaha, Stephen Junkins, Timothy R. Bauer, Varghese George, Weiyu Chen
-
Patent number: 11636174Abstract: Described herein is an accelerator device including a host interface, a fabric interconnect coupled with the host interface, and one or more hardware tiles coupled with the fabric interconnect, the one or more hardware tiles including sparse matrix multiply acceleration hardware including a systolic array with feedback inputs.Type: GrantFiled: November 16, 2021Date of Patent: April 25, 2023Assignee: Intel CorporationInventors: Subramaniam Maiyuran, Jorge Parra, Supratim Pal, Ashutosh Garg, Shubra Marwaha, Chandra Gurram, Darin Starkey, Durgesh Borkar, Varghese George
-
Publication number: 20220398375Abstract: A system and method for transferring annotations associated with a media file. An annotation associated with a media file is indexed to a first instance of that media file. By comparing features of the two instances, a mapping is created between the first instance of the media file and a second instance of the media file. The annotation can be indexed to the second instance using the mapping between the first and second instances. The annotation can be processed (displayed, stored, or modified) based on the index to the second instance.Type: ApplicationFiled: August 22, 2022Publication date: December 15, 2022Inventors: Mayur Datar, Ashutosh Garg, Vibhu Mittal
-
Publication number: 20220365901Abstract: Described herein is a graphics processing unit (GPU) comprising a first processing cluster to perform parallel processing operations, the parallel processing operations including a ray tracing operation and a matrix multiply operation; and a second processing cluster coupled to the first processing cluster, wherein the first processing cluster includes a floating-point unit to perform floating point operations, the floating-point unit is configured to process an instruction using a bfloat16 (BF16) format with a multiplier to multiply second and third source operands while an accumulator adds a first source operand with output from the multiplier.Type: ApplicationFiled: May 27, 2022Publication date: November 17, 2022Applicant: Intel CorporationInventors: Subramaniam Maiyuran, Shubra Marwaha, Ashutosh Garg, Supratim Pal, Jorge Parra, Chandra Gurram, Varghese George, Darin Starkey, Guei-Yuan Lueh
-
Publication number: 20220326953Abstract: Embodiments described herein provide for an instruction and associated logic to enable a vector multiply add instructions with automatic zero skipping for sparse input. One embodiment provides for a general-purpose graphics processor comprising logic to perform operations comprising fetching a hardware macro instruction having a predicate mask, a repeat count, and a set of initial operands, where the initial operands include a destination operand and multiple source operands. The hardware macro instruction is configured to perform one or more multiply/add operations on input data associated with a set of matrices.Type: ApplicationFiled: April 18, 2022Publication date: October 13, 2022Applicant: Intel CorporationInventors: Supratim Pal, Sasikanth Avancha, Ishwar Bhati, Wei-Yu Chen, Dipankar Das, Ashutosh Garg, Chandra S. Gurram, Junjie Gu, Guei-Yuan Lueh, Subramaniam Maiyuran, Jorge E. Parra, Sudarshan Srinivasan, Varghese George
-
Patent number: 11423213Abstract: A system and method for transferring annotations associated with a media file. An annotation associated with a media file is indexed to a first instance of that media file. By comparing features of the two instances, a mapping is created between the first instance of the media file and a second instance of the media file. The annotation can be indexed to the second instance using the mapping between the first and second instances. The annotation can be processed (displayed, stored, or modified) based on the index to the second instance.Type: GrantFiled: November 30, 2020Date of Patent: August 23, 2022Assignee: Google LLCInventors: Mayur Datar, Ashutosh Garg, Vibhu Mittal
-
Publication number: 20220261949Abstract: Examples described herein relate to a software and hardware optimization that manages scenarios where a write operation to a register is less than an entirety of the register. A compiler detects instructions that make partial writes to the same register, groups such instructions, and provides hints to hardware of the partial write. The execution unit combines the output data for grouped instructions and updates the destination register as single write instead of multiple separate partial writes.Type: ApplicationFiled: May 2, 2022Publication date: August 18, 2022Inventors: Chandra S. GURRAM, Gang Y. CHEN, Subramaniam MAIYURAN, Supratim PAL, Ashutosh GARG, Jorge E. PARRA, Darin M. STARKEY, Guei-Yuan LUEH, Wei-Yu CHEN
-
Patent number: 11416580Abstract: An apparatus to facilitate matrix multiplication operations. The apparatus comprises multiplication hardware to operate in a dot product mode, wherein a multiplication stage included in the multiplication hardware is configured as a dot product of a number of bit vectors (N) to perform N×N multiplication operations on a plurality of multiplicands and perform addition operations on results of the N×N multiplication operations.Type: GrantFiled: November 13, 2019Date of Patent: August 16, 2022Assignee: Intel CorporationInventors: Nevin Mathew, Shubra Marwaha, Ashutosh Garg