Name
Title | ||
|---|---|---|
NV-BNN: An Accurate Deep Convolutional Neural Network Based on Binary STT-MRAM for Adaptive AI Edge |
Abstract | ||
|---|---|---|
Binary STT-MRAM is a highly anticipated embedded nonvolatile memory technology in advanced logic nodes < 28 nm. How to enable its in-memory computing (IMC) capability is critical for enhancing AI Edge. Based on the soon-available STT-MRAM, we report the first binary deep convolutional neural network (NV-BNN) capable of both local and remote learning. Exploiting intrinsic cumulative switching probability, accurate online training of CIFAR-10 color images (~ 90%) is realized using a relaxed endurance spec (switching ≤ 20 times) and hybrid digital/IMC design. For offline training, the accuracy loss due to imprecise weight placement can be mitigated using a rapid non-iterative training-with-noise and fine-tuning scheme.
|
Year | DOI | Venue |
|---|---|---|
2019 | 10.1145/3316781.3317872 | Proceedings of the 56th Annual Design Automation Conference 2019 |
Keywords | Field | DocType |
binary STT-MRAM,adaptive AI Edge,nonvolatile memory technology,advanced logic nodes,in-memory computing capability,binary deep convolutional neural network,NV-BNN,STT-MRAM,cumulative switching probability,remote learning,local learning | Computer science,Convolutional neural network,Electronic engineering,Magnetoresistive random-access memory,Remote learning,Spec#,Computer engineering,Binary number | Conference |
ISSN | ISBN | Citations |
0738-100X | 978-1-4503-6725-7 | 1 |
PageRank | References | Authors |
0.48 | 5 | 11 |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Chih-Cheng Chang | 1 | 20 | 2.77 |
| Ming-Hung Wu | 2 | 2 | 1.23 |
| Jia-Wei Lin | 3 | 15 | 4.49 |
| Chun-Hsien Li | 4 | 1 | 1.16 |
| vivek parmar | 5 | 8 | 5.42 |
| Heng-Yuan Lee | 6 | 228 | 20.66 |
| Jeng-Hua Wei | 7 | 1 | 0.82 |
| Shyh-Shyuan Sheu | 8 | 211 | 23.45 |
| manan suri | 9 | 10 | 7.84 |
| Tian-Sheuan Chang | 10 | 712 | 69.10 |
| Tuo-Hung Hou | 11 | 23 | 4.92 |