Patents by Inventor Alberto Delmas Lascorz
Alberto Delmas Lascorz 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: 11928566Abstract: There is provided a system and method for compression and decompression of a data stream used by machine learning networks. The method including: encoding each value in the data stream, including: determining a mapping to one of a plurality of non-overlapping ranges, each value encoded as a symbol representative of the range and a corresponding offset; and arithmetically coding the symbol using a probability count; storing a compressed data stream including the arithmetically coded symbols and the corresponding offsets; and decoding the compressed data stream with arithmetic decoding using the probability count, the arithmetic decoded symbols use the offset bits to arrive at a decoded data stream; and communicating the decoded data stream for use by the machine learning networks.Type: GrantFiled: January 11, 2023Date of Patent: March 12, 2024Inventors: Alberto Delmas Lascorz, Andreas Moshovos
-
Publication number: 20240005213Abstract: There is provided a system and method for compression and decompression of a data stream used by machine learning networks. The method including: encoding each value in the data stream, including: determining a mapping to one of a plurality of non-overlapping ranges, each value encoded as a symbol representative of the range and a corresponding offset; and arithmetically coding the symbol using a probability count; storing a compressed data stream including the arithmetically coded symbols and the corresponding offsets; and decoding the compressed data stream with arithmetic decoding using the probability count, the arithmetic decoded symbols use the offset bits to arrive at a decoded data stream; and communicating the decoded data stream for use by the machine learning networks.Type: ApplicationFiled: September 14, 2023Publication date: January 4, 2024Inventors: Alberto DELMAS LASCORZ, Andreas MOSHOVOS
-
Publication number: 20230267376Abstract: There is provided a system and method for compression and decompression of a data stream used by machine learning networks. The method including: encoding each value in the data stream, including: determining a mapping to one of a plurality of non-overlapping ranges, each value encoded as a symbol representative of the range and a corresponding offset; and arithmetically coding the symbol using a probability count; storing a compressed data stream including the arithmetically coded symbols and the corresponding offsets; and decoding the compressed data stream with arithmetic decoding using the probability count, the arithmetic decoded symbols use the offset bits to arrive at a decoded data stream; and communicating the decoded data stream for use by the machine learning networks.Type: ApplicationFiled: January 11, 2023Publication date: August 24, 2023Inventors: Alberto DELMAS LASCORZ, Andreas Moshovos
-
Publication number: 20230186065Abstract: A system for bit-serial computation in a neural network is described. The system may be embodied on an integrated circuit and include one or more bit-serial tiles for performing bit-serial computations in which each bit-serial tile receives input neurons and synapses, and communicates output neurons. Also included is an activation memory for storing the neurons and a dispatcher. The dispatcher reads neurons and synapses from memory and communicates either the neurons or the synapses bit-serially to the one or more bit-serial tiles. The other of the neurons or the synapses are communicated bit-parallelly to the one or more bit-serial tiles, or according to a further embodiment, may also be communicated bit-serially to the one or more bit-serial tiles.Type: ApplicationFiled: February 10, 2023Publication date: June 15, 2023Applicant: Samsung Electronics Co., Ltd.Inventors: Patrick Judd, Jorge Albericio, Alberto Delmas Lascorz, Andreas Moshovos, Sayeh Sharify
-
Publication number: 20230131251Abstract: A system and method for memory compression for deep learning networks. The method includes: compacting an input data stream by identifying a bit width necessary to accommodate the value from the input data stream with the highest magnitude; storing a least significant bits of the input data stream in a first memory store, the number of bits equal to the bit width, wherein if the value requires more bits than those currently left unused in the first memory store, the remaining bits are written into a second memory store; and outputting the value of the first memory store, as a consecutive part of a compressed data stream, with an associated width of the data in the first memory store when the first memory store becomes full and copying the value of the second memory store to the first memory store; and decompressing the compressed data stream.Type: ApplicationFiled: November 10, 2022Publication date: April 27, 2023Inventors: Isak EDO VIVANCOS, Andreas MOSHOVOS, Sayeh SHARIFYMOGHADDAM, Alberto DELMAS LASCORZ
-
Patent number: 11610100Abstract: A system for bit-serial computation in a neural network is described. The system may be embodied on an integrated circuit and include one or more bit-serial tiles for performing bit-serial computations in which each bit-serial tile receives input neurons and synapses, and communicates output neurons. Also included is an activation memory for storing the neurons and a dispatcher and a reducer. The dispatcher reads neurons and synapses from memory and communicates either the neurons or the synapses bit-serially to the one or more bit-serial tiles. The other of the neurons or the synapses are communicated bit-parallelly to the one or more bit-serial tiles, or according to a further embodiment, may also be communicated bit-serially to the one or more bit-serial tiles. The reducer receives the output neurons from the one or more tiles, and communicates the output neurons to the activation memory.Type: GrantFiled: July 7, 2019Date of Patent: March 21, 2023Assignee: Samsung Electronics Co., Ltd.Inventors: Patrick Judd, Jorge Albericio, Alberto Delmas Lascorz, Andreas Moshovos, Sayeh Sharifymoghaddam
-
Publication number: 20220327367Abstract: Described is a system, integrated circuit and method for reducing ineffectual computations in the processing of layers in a neural network. One or more tiles perform computations where each tile receives input neurons, offsets and synapses, and where each input neuron has an associated offset. Each tile generates output neurons, and there is also an activation memory for storing neurons in communication with the tiles via a dispatcher and an encoder. The dispatcher reads neurons from the activation memory and communicates the neurons to the tiles and reads synapses from a memory and communicates the synapses to the tiles. The encoder receives the output neurons from the tiles, encodes them and communicates the output neurons to the activation memory. The offsets are processed by the tiles in order to perform computations only on non-zero neurons. Optionally, synapses may be similarly processed to skip ineffectual operations.Type: ApplicationFiled: June 22, 2022Publication date: October 13, 2022Applicant: Samsung Electronics Co., Ltd.Inventors: Patrick Judd, Jorge Albericio, Alberto Delmas Lascorz, Andreas Moshovos, Sayeh Sharifymoghaddam
-
Patent number: 11423289Abstract: Described is a system, integrated circuit and method for reducing ineffectual computations in the processing of layers in a neural network. One or more tiles perform computations where each tile receives input neurons, offsets and synapses, and where each input neuron has an associated offset. Each tile generates output neurons, and there is also an activation memory for storing neurons in communication with the tiles via a dispatcher and an encoder. The dispatcher reads neurons from the activation memory and communicates the neurons to the tiles and reads synapses from a memory and communicates the synapses to the tiles. The encoder receives the output neurons from the tiles, encodes them and communicates the output neurons to the activation memory. The offsets are processed by the tiles in order to perform computations only on non-zero neurons. Optionally, synapses may be similarly processed to skip ineffectual operations.Type: GrantFiled: June 14, 2017Date of Patent: August 23, 2022Assignee: Samsung Electronics Co., Ltd.Inventors: Patrick Judd, Jorge Albericio, Alberto Delmas Lascorz, Andreas Moshovos, Sayeh Sharifymoghaddam
-
Publication number: 20210004668Abstract: Described is a neural network accelerator tile for exploiting input sparsity. The tile includes a weight memory to supply each weight lane with a weight and a weight selection metadata, an activation selection unit to receive a set of input activation values and rearrange the set of input activation values to supply each activation lane with a set of rearranged activation values, a set of multiplexers including at least one multiplexer per pair of activation and weight lanes, where each multiplexer is configured to select a combination activation value for the activation lane from the activation lane set of rearranged activation values based on the weight lane weight selection metadata, and a set of combination units including at least one combination unit per multiplexer, where each combination unit is configured to combine the activation lane combination value with the weight lane weight to output a weight lane product.Type: ApplicationFiled: February 15, 2019Publication date: January 7, 2021Inventors: Andreas Moshovos, Alberto Delmas Lascorz, Zisis Poulos, Dylan Malone Stuart, Patrick Judd, Sayeh Sharify, Mostafa Mahmoud, Milos Nikolic, Kevin Chong Man Siu, Jorge Albericio
-
Publication number: 20200125931Abstract: A system for bit-serial computation in a neural network is described. The system may be embodied on an integrated circuit and include one or more bit-serial tiles for performing bit-serial computations in which each bit-serial tile receives input neurons and synapses, and communicates output neurons. Also included is an activation memory for storing the neurons and a dispatcher and a reducer. The dispatcher reads neurons and synapses from memory and communicates either the neurons or the synapses bit-serially to the one or more bit-serial tiles. The other of the neurons or the synapses are communicated bit-parallelly to the one or more bit-serial tiles, or according to a further embodiment, may also be communicated bit-serially to the one or more bit-serial tiles. The reducer receives the output neurons from the one or more tiles, and communicates the output neurons to the activation memory.Type: ApplicationFiled: July 7, 2019Publication date: April 23, 2020Inventors: Patrick Judd, Jorge Albericio, Alberto Delmas Lascorz, Andreas Moshovos, Sayeh Sharify
-
Patent number: 10387771Abstract: A system for bit-serial computation in a neural network is described. The system may be embodied on an integrated circuit and include one or more bit-serial tiles for performing bit-serial computations in which each bit-serial tile receives input neurons and synapses, and communicates output neurons. Also included is an activation memory for storing the neurons and a dispatcher and a reducer. The dispatcher reads neurons and synapses from memory and communicates either the neurons or the synapses bit-serially to the one or more bit-serial tiles. The other of the neurons or the synapses are communicated bit-parallelly to the one or more bit-serial tiles, or according to a further embodiment, may also be communicated bit-serially to the one or more bit-serial tiles. The reducer receives the output neurons from the one or more tiles, and communicates the output neurons to the activation memory.Type: GrantFiled: May 26, 2017Date of Patent: August 20, 2019Inventors: Patrick Judd, Jorge Albericio, Alberto Delmas Lascorz, Andreas Moshovos, Sayeh Sharify
-
Publication number: 20190205740Abstract: Described is a system, integrated circuit and method for reducing ineffectual computations in the processing of layers in a neural network. One or more tiles perform computations where each tile receives input neurons, offsets and synapses, and where each input neuron has an associated offset. Each tile generates output neurons, and there is also an activation memory for storing neurons in communication with the tiles via a dispatcher and an encoder. The dispatcher reads neurons from the activation memory and communicates the neurons to the tiles and reads synapses from a memory and communicates the synapses to the tiles. The encoder receives the output neurons from the tiles, encodes them and communicates the output neurons to the activation memory. The offsets are processed by the tiles in order to perform computations only on non-zero neurons. Optionally, synapses may be similarly processed to skip ineffectual operations.Type: ApplicationFiled: June 14, 2017Publication date: July 4, 2019Applicant: THE GOVERNING COUNCIL OF THE UNIVERSITY OF TORONTOInventors: Patrick Judd, Jorge Albericio, Alberto Delmas Lascorz, Andreas Moshovos, Sayeh Sharify
-
Publication number: 20170357891Abstract: A system for bit-serial computation in a neural network is described. The system may be embodied on an integrated circuit and include one or more bit-serial tiles for performing bit-serial computations in which each bit-serial tile receives input neurons and synapses, and communicates output neurons. Also included is an activation memory for storing the neurons and a dispatcher and a reducer. The dispatcher reads neurons and synapses from memory and communicates either the neurons or the synapses bit-serially to the one or more bit-serial tiles. The other of the neurons or the synapses are communicated bit-parallelly to the one or more bit-serial tiles, or according to a further embodiment, may also be communicated bit-serially to the one or more bit-serial tiles. The reducer receives the output neurons from the one or more tiles, and communicates the output neurons to the activation memory.Type: ApplicationFiled: May 26, 2017Publication date: December 14, 2017Applicant: THE GOVERNING COUNCIL OF THE UNIVERSITY OF TORONTOInventors: Patrick Judd, Jorge Albericio, Alberto Delmas Lascorz, Andreas Moshovos, Sayeh Sharify