Abstract
Abstract: State-of-the-art GPU chips are designed to deliver extreme throughput for graphics as well as for data-parallel general purpose computing workloads (GPGPU computing). Unlike graphics computing, GPGPU computing requires highly reliable operation. The performance-oriented design of GPUs requires to jointly evaluate the vulnerability of GPU workloads to soft-errors with the performance of GPU chips. We briefly present a summary of the findings of an extensive study aiming at the evaluation of the reliability of four GPU architectures and corresponding chips, orrelating them with the performance of the workloads.
Details
- BIBTEX:
@INPROCEEDINGS{7975280, author={A. Vallero and S. Di Carlo and S. Tselonis and D. Gizopoulos}, booktitle={2017 IEEE International Symposium on Performance Analysis of Systems and Software (ISPASS)}, title={Microarchitecture level reliability comparison of modern GPU designs: First findings}, year={2017}, pages={129-130}, keywords={Benchmark testing;Computer architecture;Graphics processing units;Hardware;High definition video;Measurement;Reliability;GPGPU;fault injection;microarchitecture;performance;reliability;simulator;throughput}, doi={10.1109/ISPASS.2017.7975280}, month={April},}
- 10.1109/ISPASS.2017.7975280
- KEYWORDS: Benchmark testing;Computer architecture;Graphics processing units;Hardware;High definition video;Measurement;Reliability;GPGPU;fault injection;microarchitecture;performance;reliability;simulator;throughput