This work was partially supported by grants from Intel Supercomputing Systems Division and the Oregon Ad- vanced Computing Institute (OACIS).
Parallel computers -- Design, Computer network architectures, Distributed operating systems (Computers)
Rapid advances in hardware technology have led to wide diversity in parallel computer architectures. This diversity makes it difficult to evaluate or compare the performance of different parallel computers. Existing benchmarks tend either to be too architecture-specific, or too high-level. Both problems can result in benchmarks that not only provide insufficient information on the performance characteristics of the computer being tested, but are also difficult to port. New benchmarking approaches are needed for new architectural classes, particularly distributed-memory, message-passing computers. This paper focuses on benchmarking distributed-memory message-passing computers. A synthetic benchmark called CoMet (COmmunication METrics), is presented. CoMet is based on common communication patterns found in parallel scientific algorithms. This paper presents the CoMet design, and describes an implementation of CoMet on the Intel iPSC/860. CoMet is freely available by anonymous FTP from Oregon Graduate Institute.
Ganapati, Nalini, Steve W. Otto, and Jonathan Walpole. CoMet: A Synthetic Benchmark for Message-Passing Architectures. Tech. Rep. 94-019, Oregon Graduate Institute of Science and Technology, USA, 1994.
Oregon Graduate Institute of Science & Technology Technical Report CSE-94-019, February 20, 1994.