You are in : Home » Results & submissions » Select UOA » 11 - Computer Science and Informatics » View submission: University of Leeds » Outputs » Detail

Output details

11 - Computer Science and Informatics

University of Leeds

Return to search Previous output Next output

Output 21 of 95 in the submission

Output title

Causality of Optimized Haskell: What is burning our cycles?

Type

E - Conference contribution

DOI

10.1145/2503778.2503788

Name of conference/published proceedings

Proceedings of the 2013 ACM SIGPLAN symposium on Haskell

Volume number

Issue number

First page of article

141

ISSN of proceedings

Year of publication

2013

URL

Number of additional authors

Additional information

<08>GHC is the de facto standard implementation of Haskell, spearheading adoption of Haskell by numerous commercial and scientific users. Early work on Haskell performance profiling is limited by GHC's aggressive code optimization strategies: high-performance code couldn't be profiled accurately. Our work, funded by a Microsoft PhD Scholarship (Wortmann) supported by the GHC team at MSR Cambridge (Peyton Jone, Marlow), reports a breakthrough:a systematic theory for tracing runtime costs through optimization, and full implementation within the GHC compiler. The work promises to significantly extend the reach and utility of Haskell; profiling components are beginning to be merged into GHC's master release.

Interdisciplinary

Cross-referral requested

Research group

None

Citation count

Proposed double-weighted

Double-weighted statement

Reserve for a double-weighted output

Non-English

English abstract