This paper presents our experience mapping OpenMP parallel programming model to the IBM Cyclops-64 (C64) architecture. The C64 employs a many-core-on-a-chip design that integrates...
' The coming of giga-bit networks makes possible the realization ofa single nationwide virtual computer comprised of a variety of geographically distributed highpe6ormance mac...
The underlying model of distributed systems is that of loosely coupled components r running in parallel and communicating by message passing. Description, construction and evoluti...
Hardware compilation techniques which use highlevel programming languages to describe and synthesize hardware are gaining popularity. They are especially useful for reconfigurable...
Lack of effective performance-evaluation environments is a major barrier to the broader use of high performance computing. Conventional performance environments are based on profi...