How Are We Doing? An Efficiency Measure for Shared, Heterogeneous Systems

摘要

The utilization of networked, shared, heterogeneous workstations as an inexpensive parallel computation platform is an appealing concept that has received quite a bit of attention in the literature. However, most efficiency measures for parallel computation are oriented towards the use of tightly-coupled, dedicated, homogeneous processors. We propose a new efficiency measure for parallel computations executing on networks of workstations and illustrate this efficiency measure on an example computation.