Heterogeneous SuperConcurrent Parallelism (HeSCP).

摘要

HeSCP (Heterogeneous SuperConcurrent Parallelism, pronounced 'hee-skip') is the tuned use of diverse processors to solve distinct computational needs, in which codes or code portions are directed to the processor(s) best suited for their execution. Implicit in this definition is the idea that different tasks or different portions of these tasks indicate different architectural requirements or different balances of requirements. The processors themselves are generally parallel or other HPC machines. Typically the processors are used concurrently and it is often the case that tools and methodologies used for parallel and vector processing, e.g., code profiling, flow analysis, data dependencies, etc., can be extended to the larger, meta-parallel processor. The super in SuperConcurrent refers to the traditional supercomputing aims of HeSCP, as well as the heterogeneous concurrency superimposed on the various parallel processors, and finally the aim of superlinear performance from the tuned orchestration of the heterogeneous components.