Automatic Matching of Legacy Code to Heterogeneous APIs: An Idiomatic Approach
- Submitting institution
-
University of Glasgow
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 11-04920
- Type
- E - Conference contribution
- DOI
-
10.1145/3173162.3173182
- Title of conference / published proceedings
- 23rd ACM International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS'18)
- First page
- 139
- Volume
- -
- Issue
- -
- ISSN
- 0362-1340
- Open access status
- Compliant
- Month of publication
- March
- Year of publication
- 2018
- URL
-
http://eprints.gla.ac.uk/156406/
- Supplementary information
-
-
- Request cross-referral to
- -
- Output has been delayed by COVID-19
- No
- COVID-19 affected output statement
- -
- Forensic science
- No
- Criminology
- No
- Interdisciplinary
- No
- Number of additional authors
-
5
- Research group(s)
-
-
- Citation count
- 3
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- ORIGINALITY: This paper presents an innovative approach to accelerate legacy applications on which all traditional auto-parallelisation techniques fail. The work achieves speedups of up to 200x over legacy code where alternative approaches fail. SIGNIFICANCE: Published in the premier conference for multidisciplinary systems research with 15% acceptance rate and only 15% of accepted papers from authors from European institutions. It has hence been recognised with a HiPEAC paper award. RIGOUR: The work defines a formal language to detect computational idioms in the LLVM compiler internal representation, and has been evaluated experimentally on three different hardware architectures and multiple benchmark suites.
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -