DARKSIDE: A Heterogeneous RISC-V Compute Cluster for Extreme-Edge On-Chip DNN Inference and Training
Open Access
- 27 September 2022
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Open Journal of the Solid-State Circuits Society
- Vol. 2, 231-243
Abstract
On-chip DNN inference and training at the Extreme-Edge (TinyML) impose strict latency, throughput, accuracy and flexibility requirements. Heterogeneous clusters are promising solutions to meet the challenge, combining the flexibility of DSP-enhanced cores with the performance and energy boost of dedicated accelerators. We present Darkside, a System-on-Chip with a heterogeneous cluster of 8 RISC-V cores enhanced with 2-b to 32-b mixed-precision integer arithmetic. To boost performance and efficiency on key compute-intensive Deep Neural Network (DNN) kernels, the cluster is enriched with three digital accelerators: a specialized engine for low-data-reuse depthwise convolution kernels (up to 30 MAC/cycle); a minimal overhead datamover to marshal 1-b to 32-b data on-the-fly; a 16-b floating point Tensor Product Engine (TPE) for tiled matrix-multiplication acceleration. Darkside is implemented in 65nm CMOS technology. The cluster achieves a peak integer performance of 65 GOPS and a peak efficiency of 835 GOPS/W when working on 2-b integer DNN kernels. When targeting floating-point tensor operations, the TPE provides up to 18.2 GFLOPS of performance or 300 GFLOPS/W of efficiency – enough to enable on-chip floating-point training at competitive speed coupled with ultra-low power quantized inference.Keywords
All Related Versions
Funding Information
- EU Horizon 2020 Research and Innovation through the European Pilot Project (101034126)
- WiPLASH (863337)
- ECSEL Horizon 2020 project AI4DI (826060)
- GreenWaves Technologies
This publication has 26 references indexed in Scilit:
- Eyeriss v2: A Flexible Accelerator for Emerging Deep Neural Networks on Mobile DevicesIEEE Journal on Emerging and Selected Topics in Circuits and Systems, 2019
- NEURA gheACM Transactions on Reconfigurable Technology and Systems, 2018
- Bird Sound Recognition Using a Convolutional Neural Network2018
- MobileNetV2: Inverted Residuals and Linear Bottlenecks2018
- Quantization and Training of Neural Networks for Efficient Integer-Arithmetic-Only Inference2018
- A transprecision floating-point platform for ultra-low power computing2018
- UNPU: A 50.6TOPS/W unified deep neural network accelerator with 1b-to-16b fully-variable weight bit-precision2018
- 14.1 A 2.9TOPS/W deep convolutional neural network SoC in FD-SOI 28nm for intelligent embedded systems2017
- Near-Threshold RISC-V Core With DSP Extensions for Scalable IoT Endpoint DevicesIEEE Transactions on Very Large Scale Integration (VLSI) Systems, 2017
- 14.5 Envision: A 0.26-to-10TOPS/W subword-parallel dynamic-voltage-accuracy-frequency-scalable Convolutional Neural Network processor in 28nm FDSOI2017