Title | ||
---|---|---|
ParaPIM - a parallel processing-in-memory accelerator for binary-weight deep neural networks. |
Abstract | ||
---|---|---|
Recent algorithmic progression has brought competitive classification accuracy despite constraining neural networks to binary weights (+1/-1). These findings show remarkable optimization opportunities to eliminate the need for computationally-intensive multiplications, reducing memory access and storage. In this paper, we present ParaPIM architecture, which transforms current Spin Orbit Torque Magnetic Random Access Memory (SOT-MRAM) sub-arrays to massively parallel computational units capable of running inferences for Binary-Weight Deep Neural Networks (BWNNs). ParaPIM's in-situ computing architecture can be leveraged to greatly reduce energy consumption dealing with convolutional layers, accelerate BWNNs inference, eliminate unnecessary off-chip accesses and provide ultra-high internal bandwidth. The device-to-architecture co-simulation results indicate ~4x higher energy efficiency and 7.3x speedup over recent processing-in-DRAM acceleration, or roughly 5x higher energy-efficiency and 20.5x speedup over recent ASIC approaches, while maintaining inference accuracy comparable to baseline designs.
|
Year | DOI | Venue |
---|---|---|
2019 | 10.1145/3287624.3287644 | ASP-DAC |
Field | DocType | Citations |
Computer science,Efficient energy use,Massively parallel,Parallel computing,Application-specific integrated circuit,Real-time computing,Bandwidth (signal processing),Artificial neural network,Energy consumption,Speedup,Random access | Conference | 6 |
PageRank | References | Authors |
0.39 | 11 | 3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Shaahin Angizi | 1 | 221 | 26.13 |
Zhezhi He | 2 | 136 | 25.37 |
Deliang Fan | 3 | 375 | 53.66 |