Compared with existing FPGA-based accelerator on the same network architecture, SkeletonGCN can achieve up to 11.3x speedup while maintaining the same training ...
We propose an FPGA-based GCN accelerator, named SkeletonGCN, along with multiple software-hardware co-optimizations to improve training efficiency.
To this end, we propose SkeletonGCN, a simple yet effec- tive FPGA-based accelerator for GCN training with both algo- rithm and hardware optimizations. We first ...
The authors present the work as a solution that optimizes data representation, simplifies operations, and uses a unified hardware architecture to achieve ...
Dec 31, 2019 · A novel accelerator for training GCNs on CPU-FPGA heterogeneous systems, by incorporating multiple algorithm-architecture co-optimizations ...
Kun Wang's research works | University of California, Los ...
www.researchgate.net › Kun-Wang-2058...
Additionally, SkeletonGCN features a data distribution module to control the data flow required by the MACC array, enabling efficient row data reads from the ...
Skeletongcn: a simple yet effective accelerator for gcn training. C Wu, Z Tao, K Wang, L He. 2022 32nd International Conference on Field-Programmable Logic and ...
An FPGA-based GCN accelerator, named SkeletonGCN, including multiple software-hardware co-optimizations to improve training efficiency. Low Precision Floating- ...
Nov 11, 2024 · This survey overviews recent Graph Convolutional Networks (GCN) advancements, highlighting their growing significance across various tasks and applications.
LW-GCN is proposed, a lightweight FPGA-based accelerator with a software-hardware co-designed process to tackle irregularity in computation and memory ...