I/O and computation overlap on SIMD systolic arrays

A mechanism for overlapped I/O management operations and computation on asimd linear systolic array is presented. This mechanism is based on two synchronized controllers allowing a speed-up factor of 2 oversimd machines without overlapped facility. Code generation is achieved using the C-stolic language, specifically designed for the architectural features of overlappedsimd systolic arrays.

[1]  W. Daniel Hillis,et al.  Data parallel algorithms , 1986, CACM.

[2]  T. L. Chang,et al.  Programmable Systolic Arrays , 1982, COMPCON.

[3]  Howard Jay Siegel,et al.  Modeling Overlapped Operation between the Control Unit and Processing Elements in an SIMD Machine , 1991, J. Parallel Distributed Comput..

[4]  H. T. Kung.  Why systolic architectures? , 1982, Computer.

[5]  H. T. Kung,et al.  The Warp Computer: Architecture, Implementation, and Performance , 1987, IEEE Transactions on Computers.

[6]  Shekhar Y. Borkar,et al.  iWarp: an integrated solution to high-speed parallel computing , 1988, Proceedings. SUPERCOMPUTING '88.

[7]  D. Lavenier,et al.  Experience in the design of parallel processor arrays , 1992 .

[8]  Patrice Frison,et al.  Designing specific systolic arrays with the API15C chip , 1990, [1990] Proceedings of the International Conference on Application Specific Array Processors.

[9]  Dominique Lavenier,et al.  RELACS for systolic programming , 1993, Proceedings of International Conference on Application Specific Array Processors (ASAP '93).