Issued Patents All Time
Showing 25 most recent of 32 patents
| Patent # | Title | Co-Inventors | Date |
|---|---|---|---|
| 12254061 | Apparatuses and methods to accelerate matrix multiplication | Maciej Urbanski, Brian J. Hickmann, Michael Rotzin, Andrew Yang, Brian S. Morris +1 more | 2025-03-18 |
| 12205035 | Artificial neural network training using flexible floating point tensors | Andrew Yang, Brian S. Morris | 2025-01-21 |
| 11829441 | Device and method for flexibly summing matrix values | Thomas Mark Ulrich, Ehsan Khish Ardestani Zadeh | 2023-11-28 |
| 11763131 | Systems and methods for reducing power consumption of convolution operations for artificial neural networks | — | 2023-09-19 |
| 11699081 | Systems and methods for reducing data movement during convolution operations in artificial neural networks | Ehsan Khish Ardestani Zadeh, Martin Schatz, Yuchen Hao, Abdulkadir Utku Diril, Rakesh Komuravelli | 2023-07-11 |
| 11688032 | Three-dimensional convolution pipeline with memory organizer unit | Dheevatsa Mudigere, Abdulkadir Utku Diril | 2023-06-27 |
| 11663043 | High bandwidth memory system with dynamically programmable distribution scheme | Abdulkadir Utku Diril, Olivia Wu, Anup Ramesh Kadkol, Aravind Kalaiah, Pankaj Kansal | 2023-05-30 |
| 11657252 | Point to point connected processing elements with data joiner components | Dheevatsa Mudigere, Abdulkadir Utku Diril | 2023-05-23 |
| 11640537 | Mechanism to perform non-linear functions in a machine learning accelerator | Bharat Daga, Pradeep Janedula, Aravind Srinivasan, Bijoy Pazhanimala, Ambili Vengallur | 2023-05-02 |
| 11599181 | Systems and methods for reducing power consumption of convolution operations of artificial neural networks | Abdulkadir Utku Diril, Yuchen Hao, Thomas Mark Ulrich, Rakesh Komuravelli, Ehsan Khish Ardestani Zadeh +1 more | 2023-03-07 |
| 11580192 | Grouped convolution using point-to-point connected channel convolution engines | Rakesh Komuravelli, Abdulkadir Utku Diril, Ehsan Khish Ardestani Zadeh, Yuchen Hao, Martin Schatz +4 more | 2023-02-14 |
| 11537865 | Mapping convolution to a channel convolution engine | Rakesh Komuravelli, Abdulkadir Utku Diril, Ehsan Khish Ardestani Zadeh, Yuchen Hao, Martin Schatz +4 more | 2022-12-27 |
| 11537301 | High bandwidth memory system with distributed request broadcasting masters | Abdulkadir Utku Diril, Olivia Wu, Aravind Kalaiah, Anup Ramesh Kadkol, Pankaj Kansal | 2022-12-27 |
| 11531619 | High bandwidth memory system with crossbar switch for dynamically programmable distribution scheme | Olivia Wu, Abdulkadir Utku Diril, Aravind Kalaiah, Anup Ramesh Kadkol, Pankaj Kansal | 2022-12-20 |
| 11520853 | Mapping convolution to a partition channel convolution engine | Rakesh Komuravelli, Abdulkadir Utku Diril, Ehsan Khish Ardestani Zadeh, Yuchen Hao, Martin Schatz +4 more | 2022-12-06 |
| 11520854 | Support for different matrix multiplications by selecting adder tree intermediate results | Yuchen Hao, Ehsan Khish Ardestani Zadeh, Rakesh Komuravelli, Abdulkadir Utku Diril, Thomas Mark Ulrich | 2022-12-06 |
| 11501147 | Systems and methods for handling padding regions in convolution operations | Ehsan Khish Ardestani, Martin Schatz, Yuchen Hao, Abdulkadir Utku Diril, Rakesh Komuravelli | 2022-11-15 |
| 11481471 | Mapping convolution to a matrix processor unit | Abdulkadir Utku Diril, Dheevatsa Mudigere, Olivia Wu, Ehsan Khish Ardestani Zadeh, Yuchen Hao | 2022-10-25 |
| 11455143 | Using a low-bit-width dot product engine to sum high-bit-width numbers | Thomas Mark Ulrich, Ehsan Khish Ardestani Zadeh | 2022-09-27 |
| 11443013 | Pipelined pointwise convolution using per-channel convolution operations | Rakesh Komuravelli, Abdulkadir Utku Diril, Ehsan Khish Ardestani Zadeh, Yuchen Hao, Martin Schatz +4 more | 2022-09-13 |
| 11409838 | High throughput matrix processor with support for concurrently processing multiple matrices | Olivia Wu, Ehsan Khish Ardestani Zadeh, Abdulkadir Utku Diril, Thomas Mark Ulrich, Yuchen Hao +2 more | 2022-08-09 |
| 11379557 | Device and method for flexibly summing matrix values | Thomas Mark Ulrich, Ehsan Khish Ardestani Zadeh | 2022-07-05 |
| 11372644 | Matrix processing instruction with optional up/down sampling of matrix | Thomas Mark Ulrich, Yuchen Hao | 2022-06-28 |
| 11275560 | Hardware for floating-point arithmetic in multiple formats | Thomas Mark Ulrich, Abdulkadir Utku Diril, Zhao-Jun Wang, Rakesh Komuravelli | 2022-03-15 |
| 11188303 | Floating point multiply hardware using decomposed component numbers | Anup Ramesh Kadkol, Ehsan Khish Ardestani Zadeh, Olivia Wu, Yuchen Hao, Thomas Mark Ulrich +1 more | 2021-11-30 |