The documents distributed by this server have been provided by the contributing authors as a means to ensure timely dissemination of scholarly and technical work on a noncommercial basis. Copyright and all rights therein are maintained by the authors or by other copyright holders, notwithstanding that they have offered their works here electronically. It is understood that all persons copying this information will adhere to the terms and constraints invoked by each author's copyright. These works may not be reposted without the explicit permission of the copyright holder.
Publications of SPCL
|M. Martinasso, G. Kwasniewski, S. R. Alam, T. Schulthess, T. Hoefler:
|A PCIe Congestion-Aware Performance Model for Densely Populated Accelerator Servers
(In Proceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis (SC16), presented in Salt Lake City, Utah, pages 63:1--63:11, IEEE Press, ISBN: 978-1-4673-8815-3, Nov. 2016)
AbstractMeteoSwiss, the Swiss national weather forecast institute, has selected densely populated accelerator servers are their primary system to compute weather forecast simulation. Servers with multiple accelerator devices that are primarily connected by a PCI-Express (PCIe) network achieve a significantly higher energy efficiency. Memory transfers between accelerators in such a system are subjected to PCIe arbitration policies. In this paper, we study the impact of PCIe topology and develop a congestion-aware performance model for PCIe communication. We present an algorithm for computing penalty coefficients of every communication in a congestion graph that characterises the dynamic usage of network resources by an application. Our validation results on two different topologies of 8 GPU devices demonstrate that our model achieves an accuracy of over 97% within the PCIe network. We use the model on a weather forecast application to identify the best algorithm for its communication patterns among GPUs.
Recorded talk (best effort)