Issued Patents 2024
Showing 1–8 of 8 patents
| Patent # | Title | Co-Inventors | Date |
|---|---|---|---|
| 12135968 | Instructions to convert from FP16 to BF8 | Alexander Heinecke, Robert Valentine, Mark J. Charney, Christopher J. Hughes, Evangelos Georganas +3 more | 2024-11-05 |
| 12106210 | Scaling half-precision floating point tensors for training deep neural networks | Dipankar Das | 2024-10-01 |
| 12056489 | Apparatuses, methods, and systems for 8-bit floating-point matrix dot product instructions | Alexander Heinecke, Robert Valentine, Mark J. Charney, Christopher J. Hughes, Evangelos Georganas +3 more | 2024-08-06 |
| 12033237 | Dynamic precision management for integer deep learning primitives | Dheevatsa Mudigere, Dipankar Das, Srinivas Sridharan | 2024-07-09 |
| 12020028 | Apparatuses, methods, and systems for 8-bit floating-point matrix dot product instructions | Alexander Heinecke, Robert Valentine, Mark J. Charney, Christopher J. Hughes, Evangelos Georganas +3 more | 2024-06-25 |
| 11977885 | Utilizing structured sparsity in systolic arrays | Subramaniam Maiyuran, Jorge Parra, Ashutosh Garg, Chandra Gurram, Chunhui Mei +10 more | 2024-05-07 |
| 11900107 | Instructions for fused multiply-add operations with variable precision input operands | Dipankar Das, Mrinmay Dutta, Arun Kumar, Dheevatsa Mudigere, Abhisek KUNDU | 2024-02-13 |
| 11893490 | Incremental precision networks using residual inference and fine-grain quantization | Abhisek KUNDU, Dheevatsa Mudigere, Dipankar Das | 2024-02-06 |