Grigori Fursin
Grigori Fursinb | |
---|---|
Born | 1977 (age 39–40) |
Fields |
Computer engineering Machine learning |
Institutions |
|
Alma mater | |
Thesis | Iterative Compilation and Performance Prediction for Numerical Applications (2004 ) |
Known for | MILEPOST GCC, cTuning foundation, Collective Knowledge framework, Artifact Evaluation at IEEE/ACM conferences |
Notable awards |
|
Website fursin |
Grigori Fursin (born 1977) is a computer scientist, president of the cTuning Foundation, CTO and co-founder of dividiti. His research group created machine learning based self-tuning compiler, MILEPOST GCC,[2] considered by IBM to be the first in the world.[3] At the end of the MILEPOST project he established cTuning.org portal and the cTuning foundation bringing academia and industry together to develop self-optimizing computer systems by crowdsourcing program optimization and machine learning across diverse devices provided by volunteers.[4][5] His foundation also developed Collective Knowledge Framework to support open research and help the community share their artifacts and workflows as reusable and customizable Python components with a unified API. In 2017 Fursin and his colleagues won the Test of Time award for their CGO'07 research paper on using machine learning and performance counters to predict compiler optimizations - this annual award recognizes outstanding papers published at the ACM/IEEE International Symposium on Code Generation and Optimization (CGO) one decade earlier, whose influence is still strong today.[1] Since 2015 he leads Artifact Evaluation at several ACM and IEEE computer systems conferences to enable open, collaborative and reproducible research.
Education
Fursin received a Master of Science degree in physics and mathematics from Moscow Institute of Physics and Technology in 1999. He completed his PhD in computer science at the University of Edinburgh in 2005. While in Edinburgh, he worked on foundations of practical program autotuning and performance prediction.[6]
Notable projects
- Collective Knowledge - open-source framework to help researchers preserve and reuse their knowledge (code and data) in a form of simple and customizable Python widgets with unified JSON API, quickly prototype research workflows from shared components, crowdsource experiments and reproduce results
- CK-AI - open experimental framework powered by Collective Knowledge to enable collaborative co-design and optimization of the whole software and hardware stack of Deep learning in terms of speed, accuracy, energy, size and cost across diverse models, inputs and devices
- MILEPOST GCC - open-source technology to build machine learning based compilers
- Interactive Compilation Interface - plugin framework to expose internal features and optimization decisions of compilers for external auto tuning and learning.[4] Available in mainline GCC since version 4.5[7]
- Artifact Evaluation - collaborative validation of experimental results from published papers at the leading ACM and IEEE computer systems conferences
- cknowledge.org/repo - public repository to crowdsource program optimization across diverse devices such as mobile phones and HPC servers provided by volunteers
- cTuning foundation - non-profit research organization developing open-source tools and common methodology for collaborative and reproducible experimentation
References
- 1 2 HiPEAC info 50 (page 8) (PDF), April 2017
- ↑ Grigori Fursin, Yuriy Kashnikov, Abdul Wahid Memon, Zbigniew Chamski, Olivier Temam, Mircea Namolaru, Elad Yom-Tov, Bilha Mendelson, Ayal Zaks, Eric Courtois, Francois Bodin, Phil Barnard, Elton Ashton, Edwin Bonilla, John Thomson, Chris Williams, Michael O'Boyle. Milepost gcc: Machine learning enabled self-tuning compiler International journal of parallel programming, Volume 39, Issue 3, pp. 296-327, June 2011 (link)
- ↑ World's First Intelligent, Open Source Compiler Provides Automated Advice on Software Code Optimization, IBM press-release, June 2009 (link)
- 1 2 Grigori Fursin. Collective Tuning Initiative: automating and accelerating development and optimization of computing systems. Proceedings of the GCC Summit'09, Montreal, Canada, June 2009 (link)
- ↑ Fursin, Grigori; Abdul Memon; Christophe Guillon; Anton Lokhmotov (January 2015). Collective Mind, Part II: Towards Performance- and Cost-Aware Software Engineering as a Natural Science. Proceedings of the CPC 2016.
- ↑ Grigori Fursin. "Resume". Retrieved 2017-05-21.
- ↑ "GCC plugins". Retrieved 2017-05-30.