Improving Strong-Scaling on GPU Cluster Based on Tightly Coupled Accelerators Architecture
Toshihiro Hanawa, Hisafumi Fujii, Norihisa Fujita, Tetsuya Odajima, Kazuya Matsumoto, Yuetsu Kodama, Taisuke Boku · 2015
The Tightly Coupled Accelerators (TCA) architecture that we proposed in previous work enables direct ommunication between accelerators over nodes. In this paper, we present a proof-of-concept GPU cluster called the HA-PACS/TCA using the PEACH2 chip that we designed as an interconnection router chip based on the TCA architecture. Our system demonstrated 2.0 ?sec of latency on inter-node GPU-to-GPU communication with a PCIe Gen2 x8 by RDMA, reducing minimum latency to just 44% of the InfiniBand-QDR and MPI using GPUDirect for RDMA. Through results of Himeno benchmark tests, we demonstrated that our TCA architecture improved performance scalability with the small-sized problem by up to 61%.