Performance evaluation of acceleration of convolutional layers on OpenEdgeCGRA
Abstract: Recently, efficiently deploying deep learning solutions on the edge has received increasing attention. New platforms are emerging to support the increasing demand for flexibility and high performance. In this work, we explore the efficient mapping of convolutional layers on an open-hardware, low-power Coarse-Grain Reconfigurable Array (CGRA), namely OpenEdgeCGRA. We explore both direct implementations of convolution and solutions that transform it into a matrix multiplication through an Im2col transformation, and experiment with various tensor parallelism axes. We show that for this hardware target, direct convolution, coupled with weight parallelism reaches the best latency and energy efficiency, outperforming a CPU implementation by 3.4x and 9.9x in terms of energy and latency, respectively.
- Norman P Jouppi et al. In-datacenter performance analysis of a tensor processing unit. In Proc. of the 44th ISCA, pages 1–12, 2017.
- Yixing Li et al. A gpu-outperforming fpga accelerator architecture for binary convolutional neural networks. ACM JETC, 14:1 – 16, 2017.
- Specializing cgras for light-weight convolutional neural networks. IEEE TCAD, 41(10):3387–3399, 2021.
- Dhananjaya Wijerathne et al. Himap: Fast and scalable high-quality mapping on cgra via hierarchical abstraction. IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 41(10):3290–3303, 2021.
- Christian Heidorn et al. Design space exploration for layer-parallel execution of convolutional neural networks on cgras. In Proc. of the 23th SCOPES, pages 26–31, 2020.
- Artur Podobas et al. A survey on coarse-grained reconfigurable architectures from a performance perspective. IEEE Access, 8:146719–146743, 2020.
- Rubén Rodríguez Álvarez et al. An open-hardware coarse-grained reconfigurable array for edge computing. In Proc. of the 20th ACM CF, pages 391–392, 2023.
- X-heep: An open-source, configurable and extendible risc-v microcontroller for the exploration of ultra-low-power edge accelerators. arXiv preprint arXiv:2401.05548, 2024.
- Deep learning, volume 1. MIT Press, 2017.
- Liangzhen Lai et al. CMSIS-NN: efficient neural network kernels for arm cortex-m cpus. CoRR, abs/1801.06601, 2018.
- Angelo Garofalo et al. Pulp-nn: accelerating quantized neural networks on parallel ultra-low-power risc-v processors. Phil. Trans. of the Royal Society A, 378(2164):20190155, 2020.
- Tianqi Chen et al. Mxnet: A flexible and efficient machine learning library for heterogeneous distributed systems. CoRR, abs/1512.01274, 2015.
- Sanjay Krishnan et al. Artificial intelligence in resource-constrained and shared environments. SIGOPS Oper. Syst. Rev., 53(1):1–6, jul 2019.
- Vivienne Sze et al. Designing DNN Accelerators, chapter 5, pages 73–118. Springer Cham, 2022.
Paper Prompts
Sign up for free to create and run prompts on this paper using GPT-5.
Top Community Prompts
Collections
Sign up for free to add this paper to one or more collections.