Performance Characteristics of Intel Architecture — based Servers

Computing clusters built from standard components using Intel® processors are becoming the fastest growing choice for high-performance computing (HPC). Twice yearly, the 500 most powerful computing systems in the world are ranked on the TOP500 Supercomputer Sites Web page. In November 2002, the ranking listed 56 entries using Intel processors; by June 2003, that number reached 119. Today, three of the top 10 computing systems in the world are clusters of Intel architecture–based servers.1 Since 1995, Intel has continuously introduced more powerful processors with increased speed, faster frontside bus, and larger cache. To test the effectiveness of Intel processor–based servers in HPC environments, a DellTM team compared performance characteristics of three Intel architecture–based Dell PowerEdgeTM servers using benchmarks commonly employed in HPC. Studies have shown that small-scale symmetric multiprocessing (SMP) systems make excellent platforms for building HPC clusters.2 Thus, two-processor systems were used for the Dell tests discussed in this article. Establishing the test environment The test environment consisted of three Dell PowerEdge server models, each with a different Intel processor: Intel Pentium® III, Intel XeonTM, and Intel Itanium® 2. Figure 1 shows the specifications for each server. The software environment included the Red Hat® Linux® operating system, the MPI/Pro® Message Passing Interface (MPI) library from MPI Software Technology for running multiple processes, and various Intel compilers. Benchmarks used to compare performance among the three platforms included High-Performance Linpack (HPL), the NAS (NASA Advanced Supercomputing) Parallel Benchmarks (NPB), and a kernel benchmark called Additive Schwarz Preconditioned Conjugate Gradient (ASPCG).