GRAM
Open Access
- 9 February 2021
- journal article
- research article
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Architecture and Code Optimization
- Vol. 18 (2), 1-24
- https://doi.org/10.1145/3441830
Abstract
This article presents GRAM (GPU-based Runtime Adaption for Mixed-precision) a framework for the effective use of mixed precision arithmetic for CUDA programs. Our method provides a fine-grain tradeoff between output error and performance. It can create many variants that satisfy different accuracy requirements by assigning different groups of threads to different precision levels adaptively at runtime . To widen the range of applications that can benefit from its approximation, GRAM comes with an optional half-precision approximate math library. Using GRAM, we can trade off precision for any performance improvement of up to 540%, depending on the application and accuracy requirement.Keywords
Funding Information
- Singapore Ministry of Education (T1-251RES1818)
This publication has 39 references indexed in Scilit:
- A Survey of Techniques for Approximate ComputingACM Computing Surveys, 2016
- Approximating warps with intra-warp operand value similarityPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2016
- Mitigating the Memory Bottleneck With Approximate Load Value PredictionIEEE Design & Test, 2016
- Approximate Computing: A SurveyIEEE Design & Test, 2015
- Neural acceleration for GPU throughput processorsPublished by Association for Computing Machinery (ACM) ,2015
- Scalable-effort classifiers for energy-efficient machine learningPublished by Association for Computing Machinery (ACM) ,2015
- Approximate computing and the quest for computing efficiencyPublished by Association for Computing Machinery (ACM) ,2015
- SAGEPublished by Association for Computing Machinery (ACM) ,2013
- EnerJACM SIGPLAN Notices, 2011
- Toward efficient static analysis of finite-precision effects in DSP applications via affine arithmetic modelingPublished by Association for Computing Machinery (ACM) ,2003