A precise measurement tool for power dissipation of CUDA kernels
スポンサーリンク
概要
- 論文の詳細を見る
Power dissipation has become an important factor for evaluating application performance. To help programmers have a further understand on the power dissipation of their applications, we are investigating tools to achieve precise and easily using measurement tools. In this paper we discuss power measurements of GPU, propose an API tool which include two parts: host part and monitor part. The host part is used for measurement initial, communication with monitor machine and data process. The monitor part is used for real-time voltages and current data collection. By using this tool, programmers can get accurate power dissipation of their applications.
- 2012-03-19
著者
-
Reiji Suda
Presently With Crest Jst
-
Luo Cheng
Presently with The University of Tokyo
-
Luo Cheng
Presently With The University Of Tokyo|presently With Crest Jst
-
Kamil Rocki
Presently With The University Of Tokyo|presently With Crest Jst
関連論文
- An execution time prediction analytical model for GPU with instruction-level and thread-level parallelism awareness
- A precise measurement tool for power dissipation of CUDA kernels
- A Three-Step Performance Automatic Tuning Strategy using Statistical Model for OpenCL Implementation of Krylov Subspace Methods
- Efficient Monte Carlo Optimization with ATMathCoreLib
- Evaluation of Impact of Noise on Collective Algorithms in Repeated Computation Cycle
- The Future of Accelerator Programming: Abstraction, Performance or Can We Have Both?