Title :
Application Performance Tuning for Clusters with ccNUMA Nodes
Author :
Kayi, Abdullah ; Kornkven, Edward ; El-Ghazawi, Tarek ; Newby, Greg
Author_Institution :
Dept. of Electr. & Comput. Eng., George Washington Univ., Washington, DC
Abstract :
With the increasing trend of putting more cores inside a single chip, more clusters adapt multicore multiprocessor nodes for high-performance computing (HPC). Cache coherent non-uniform memory access architectures (ccNUMA) are becoming an increasingly popular choice for such systems. In this paper, application performance analysis is provided using a 2312 Opteron cores system based on Sun Fire servers. Performance bottlenecks are identified and some potential solutions are proposed. With the proposed performance tunings, up to 30% application performance improvement was observed. In addition, provided experimental analysis can be utilized by HPC application developers in order to better understand clusters with ccNUMA nodes and also as a guideline for the usage of such architectures for scientific computing.
Keywords :
cache storage; memory architecture; multiprocessing systems; 2312 Opteron cores system; application performance tuning; cache coherent nonuniform memory access architecture; ccNUMA nodes; high-performance computing; multicore multiprocessor nodes; Arctic; Benchmark testing; Computer architecture; Fires; High performance computing; Memory architecture; Multicore processing; Scalability; Sockets; Sun; application performance; ccNUMA; cpu affinity; high-performance computing;
Conference_Titel :
Computational Science and Engineering, 2008. CSE '08. 11th IEEE International Conference on
Conference_Location :
Sao Paulo
Print_ISBN :
978-0-7695-3193-9
DOI :
10.1109/CSE.2008.46
