Latency, Bandwidth, and Concurrent Issue Limitations in High-Performance CFD

TitleLatency, Bandwidth, and Concurrent Issue Limitations in High-Performance CFD
Publication TypeReport
Year of Publication2000
AuthorsGropp, WD, Kaushik, DK, Keyes, DE, Smith, BF
Date Published10/2000
Other NumbersANL/MCS-P850-1000
Abstract

To achieve high performance, a parallel algorithm needs to effectively
utilize the memory subsystem and minimize the communication volume
and the number of network transactions. These issues gain further importance on modern architectures, where the peak CPU performance is
increasing much more rapidly than the memory or network performance.
In this paper, we present some performance enhancing techniques that
were employed on an unstructured mesh implicit solver. Our experimental results show that this solver adapts resonably well to the high memory
and network latencies.

PDFhttp://www.mcs.anl.gov/papers/P850.pdf