Measuring Parallelism in Computation-Intensive Scientific/Engineering Applications
IEEE Transactions on Computers
Computer - Special issue on experimental research in computer architecture
Single instruction stream parallelism is greater than two
ISCA '91 Proceedings of the 18th annual international symposium on Computer architecture
On the limits of program parallelism and its smoothability
MICRO 25 Proceedings of the 25th annual international symposium on Microarchitecture
A Quantitative Approach for Architecture-Invariant Parallel Workload Characterization
PARA '96 Proceedings of the Third International Workshop on Applied Parallel Computing, Industrial Computation and Optimization
Measuring the Parallelism Available for Very Long Instruction Word Architectures
IEEE Transactions on Computers
| Hi-index | 0.00 |
In this paper, a model is presented for representing and comparing workloads, based on the way they would exercise parallel machines. This workload characterization is derived from parallel instruction centroid and parallel workload similarity. The centroid is a simple measure that aggregates average parallelism, instruction mix, and critical path length. When captured with abstracted information about communication requirements, the result is a powerful tool in understanding the requirements of workloads and their potential performance on target machines. The workload similarity is based on measuring the normalized Euclidean distance (ned) between workload centroids. It will be shown that this workload representation method outperforms comparable ones in accuracy, as well as in time and space requirements. Analysis of the NAS Parallel Benchmarks and their performance is presented to demonstrate some of the applications, such as performance prediction with good accuracy, and insight provided by this model.