Name
Abstract | ||
|---|---|---|
The demands of applications using a high-speed deep learning models at data centers are rapidly increasing. However, most of these accelerators depend on many memory accesses and DSP blocks, which cause performance bottleneck. We present a lookup table (LUT) mapping to directly map convolutional layers, mainly used in modern deep learning models. To reduce the number of LUTs, we develop a training... |
Year | DOI | Venue |
|---|---|---|
2021 | 10.1109/ISMVL51352.2021.00032 | 2021 IEEE 51st International Symposium on Multiple-Valued Logic (ISMVL) |
Keywords | DocType | ISSN |
Training,Deep learning,Data centers,Technological innovation,Solid modeling,Convolution,Tools | Conference | 0195-623X |
ISBN | Citations | PageRank |
978-1-7281-9224-6 | 0 | 0.34 |
References | Authors | |
0 | 3 | |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Naoto Soga | 1 | 0 | 0.34 |
| Ryosuke Kuramochi | 2 | 0 | 2.70 |
| Hiroki Nakahara | 3 | 155 | 37.34 |