论文信息 - Using hls4ml to Map Convolutional Neural Networks on Interconnected FPGA Devices - 字舞流文

Using hls4ml to Map Convolutional Neural Networks on Interconnected FPGA Devices

Workflow used in this paper to partition CNNs in multiple FPGAs.

Manolis Katevenis | Nikolaos Chrysos | Nikolaos Dimou | Evangelos Mageiropoulos

[1] Jason Cong,et al. Energy-Efficient CNN Implementation on a Deeply Pipelined FPGA Cluster , 2016, ISLPED.

[2] Luiz André Barroso,et al. The tail at scale , 2013, CACM.

[3] R. Siezen,et al. others , 1999, Microbial Biotechnology.

[4] Yu Wang,et al. [DL] A Survey of FPGA-based Neural Network Inference Accelerators , 2019, ACM Trans. Reconfigurable Technol. Syst..

[5] Forrest N. Iandola,et al. SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <1MB model size , 2016, ArXiv.

[6] Michael Ferdman,et al. Maximizing CNN accelerator efficiency through resource partitioning , 2016, 2017 ACM/IEEE 44th Annual International Symposium on Computer Architecture (ISCA).

[7] Song Han,et al. Fast inference of deep neural networks in FPGAs for particle physics , 2018, Journal of Instrumentation.