Grigori Fursin, PhD

Academic CV

LinkedIn    Publications    Twitter    GitHub    Contact
My interests: efficient computing | reproducibility | ML for systems | ML-based auto-tuning | self-optimizing systems | automated ML/SW/HW co-design | portability | reusable artifacts | workflow automation | micro-services | DevOps | MLCommons | MLPerf | MLOps | EdgeAI | deep tech | knowledge management | collaborative learning | online teaching | open science

I am a computer scientist, engineer, educator and business executive with an interdisciplinary background in computer engineering, machine learning, physics and electronics. I am passionate about designing efficient systems in terms of speed, energy, accuracy and various costs, bringing deep tech to the real world, teaching, enabling reproducible research and supporting open science.

My academic research (tenured research scientist at INRIA with PhD in CS from the University of Edinburgh)
  • I was among the first researchers to combine machine learning, autotuning and knowledge sharing to automate and accelerate the development of efficient software and hardware by several orders of magnitude (Google scholar);
  • developed open-source tools and started educational initiatives (ACM, Raspberry Pi foundation) to bring this research to the real world (see use cases);
  • prepared and tought M.S. course at Paris-Saclay University on using ML to co-design efficient software and hardare (self-optimizing computing systems);
  • gave 100+ invited research talks;
  • honored to receive the ACM CGO test of time award, several best papers awards and INRIA award of scientific excellence.
Project management, system design and consulting (collaboration with IBM, Intel, Arm, Synopsys, Google, Mozilla, General Motors)
  • led the development of the world's first ML-based compiler and the cTuning.org platform across 5 teams to automate and crowdsource optimization of computer systems (IBM and Fujitsu press-releases; invitation to help establish Intel Exascale Lab and lead SW/HW co-design group);
  • developed a compiler plugin framework that was added to the mainline GCC powering all Linux-based computers and helped to convert production compilers into research toolsets for machine learning;
  • developed the Collective Knowledge framework to automate and accelerate design space exploration of AI/ML/SW/HW stacks while balancing speed, accuracy, energy and costs (188K+ downloads); CK helped to automate most of MLPerf inference benchmark submissions for edge devices as mentioned by Forbes, ZDNet and EETimes;
  • co-founded an engineering company and led it to $1M+ in revenue with Fortune 50 customers using my CK technology;
Startups
  • founded and developed the cKnowledge.io platform to organize all research knowledge (AI, ML, Systems, quantum) in the form of portable CK workflows, common APIs and reusable artifacts (acquired by OctoML.ai). It helps users quickly find innovative technology, test it and adopt in production.
Community service (collaboration with ACM, MLPerf and the Raspberry Pi foundation)

Quick access:
[ Professional experience (J) ], [ Startups (C) ], [ Institution building (I) ], [ Editor (Ed) ], [ Keynotes (K) ], [ Social activities and community service (O) ], [ Examiner ], [ Expert service (E) ], [ Major research achievements (M) ], [ Public or in-house repositories of knowledge (R) ], [ Awards, prizes, and fellowships (A) ], [ Major funding (F) ], [ Education (Z) ], [ Major software and datasets (S) ], [ Hardware (H) ], [ Talks (T) ], [ Participating in program committees and reviewing ], [ Teaching and organizing courses (L) ], [ Advising/collaborating (Q) ], [ Organizing/chairing events (E) ], [ Publications (P) ]

Professional experience

# Year Job
[J1] 2021-04 - cur. VP of MLOps at OctoML.ai

[ company website ]
[J2] 2019-06 - 2021-03 Founder of cKnowledge.io (acquired by OctoML.ai) building an open knowledge repository about how to design, benchmark, optimize and use complex computational systems for AI, ML, quantum, IoT and other emerging workloads, accelerate their adoption, and reduce the technical debt

[ The Collective Knowledge portal ][ MLPerf crowd-benchmarking demo ]
[J3] 2016 - 2018 Partner in the AI/ML/SW/HW co-design project of General Motors (USA) using my CK framework with the universal autotuning workflow for automated co-design of Pareto-efficient software/hardware stacks for AI/ML workloads across diverse models, data set and platforms in terms of speed, accuracy, energy, and costs.

[ GM presentation about my Collective Knowledge technology at the Embedded Vision Summit ]
[J4] 2015-07 - 2019-03 Co-founder and CTO of dividiti (UK) pursuing my vision of self-optimizing, efficient, reliable and affordable computing from IoT to supercomputers. We offered professional services based on my Collective Knowledge technology to help our Fortune50 clients and startups automate benchmarking, optimization and co-design of emerging AI, ML and quantum workloads using my CK framework.

[ The summary of my experience: ][ Our partners ]
[J5] 2014-11 - cur. Founder, CEO and CSO of the cTuning foundation (France) developing open-source technology to automate AI, ML, computer systems and quantum R&D, and accelerate technology transfer from academia to industry to solve the world's most challenging problems!
  • Funded by EU FP7 TETRACOM project [F3]
  • Sponsored by Microsoft (access to Microsoft Azure cloud) [F1]
  • Developed Collective Knowledge software and repository [S2, R2]
  • Received HiPEAC award [A3]
  • Publications [P14]
[J6] 2014 - 2015 Partner in the EU TETRACOM project of Arm (UK) developing the open-source Collective Knowledge framework with a permissive license and validating it in practice with Arm thanks to a 6-month grant form the EU TETRACOM initiative. I was honored to receive the European technology transfer award for this project.

[ The Collective Knowledge project ][ CK framework ]
[J7] 2011-09 - 2014-10 Tenured research scientist (associate professor) at INRIA Saclay (France) directing R&D of the novel Collective Mind concept for collaborative, systematic and reproducible benchmarking, optimization and co-design of computer systems using public repository of knowledge, plugin-based auto-tuning, run-time adaptation, big data, predictive analytics, machine learning, data mining, statistical analysis, feature selection, crowdsourcing and collective intelligence
  • Funded by INRIA 4 year fellowship [A4]
  • Initiated new publication model supported by HiPEAC where experimental results (tools, data, models) are continuously shared and validated by the community [E25, E27]
  • Associated software [S3, S10, S4, S5, S6]
  • Associated publications [P17, P16, P18, P19, P20, P21, P22, P23, P24, P25]
  • Associated events [E23, E25, E26, E28, E29, E30]
  • Associated public repository of knowledge [R3]
[J8] 2010-03 - 2011-08 Co-director and the head of the application optimization group in Intel Exascale Lab (France) preparing the long-term R&D vision based on my past research while leading the team of 8 researchers, engineers and students
  • More info about this activity [I2]
[J9] 2008-01 - 2008-01 Visiting scientist in the Institute of Computing Technology of Chinese Academy Of Sciences preparing collaboration on extending cTuning and MILEPOST technology
  • Funded by ICT (China) [F8]
  • Associated publications [P23, P30, P32, P37]
[J10] 2007-09 - 2010-02 Tenured research scientist (assistant professor) at INRIA Saclay (France) leading R&D in EU MILEPOST project on building practical machine learning based research compiler (MILEPOST GCC) and public plugin-based auto-tuning framework and repository of knowledge (cTuning.org)
  • Funded by EU MILEPOST project [F9]
  • Associated job [J12]
  • Associated public repository of knowledge [R5]
[J11] 2007-09 - 2014-10 Adjunct professor at University of Paris-Sud (France) preparing and teaching my novel approach on collaborative, systematic and reproducible benchmarking, optimization and co-design of computer systems using public repository of knowledge, plugin-based auto-tuning, run-time adaptation, big data, predictive analytics, machine learning, data mining, statistical analysis, feature selection, crowdsourcing and collective intelligence
  • Promoting my new research and publication model where experimental results (tools, data, models) are continuously shared and validated by the community [P14, E25, E27]
  • Sharing all my related code and data at cTuning.org and later Collective Mind repository [R3]
  • Regularly giving invited lectures in the UK, USA, Canada, Russia and Asia
[ Sharing all notes online through cTuning foundation ][ Enabling interactive publications ]
[J12] 2006-07 - 2009-06 Technical leader of the EU FP6 035307 MILEPOST project directing development of the world's first practical machine learning based self-tuning compiler and of the first public repository of optimization knowledge to crowdsource optimization and co-design of computer systems
[J13] 2005-12 - 2007-08 Postdoctoral researcher and principal investigator at INRIA Saclay (France) preparing foundations of crowdsourcing auto-tuning combined with machine learning and public repositories of knowledge
[J14] 2004-10 - 2005-11 Visiting scientist at INRIA Saclay (France) leading R&D of a novel concept of statically enabling dynamic optimizations using multi-versioning and run-time adaptation
  • Funded by EU HiPEAC fellowship [A8]
  • Prepared base for crowdtuning
  • Associated publications [P57]
[J15] 2002-01 - 2005-11 Research associate at the University of Edinburgh (UK) leading research and developing framework and repository for program online auto-tuning, polyhedral optimization, and machine learning-based software/hardware co-design and co-optimization
[J16] 2000-02 - 2000-03 Visiting scientist at Paris South University and INRIA Saclay (France) leading development of novel memory/CPU characterization technique via semantically non-equivalent binary patching
[J17] 1999-02 - 2001-12 Research assistant (research associate since 2000) in the University of Edinburgh (UK) leading R&D in 2 workpackages in the EU MHAOTEU project on program behavior analysis and auto-tuning for HPC systems
[J18] 1999-02 - cur. Evangelist of a collaborative and reproducible research and experimentation in computer engineering
  • Developing public repositories of knowledge and common research and development tools [R8, R7, R6, R5, R3]
  • Enabling new publication model where experimental results (tools, data, models) are continuously shared and validated by the community [P14, E25, E27]
  • Established international not-for-profit cTuning foundation [O6]
[ cTuning foundation activities on collaborative and reproducible research, development and experimentation in computer engineering ][ Public wiki on reproducible research ][ My motivation, history and manifesto ]
[J19] 1998-09 - 1999-01 Programmer in the Laboratory for Computer Technologies in Teaching in Moscow Insitute of Physics and Technology (Russia) leading development of educational web-based software for undegraduate courses in computer engineering and machine learning
[J20] 1997-09 - 1999-02 Research assistant and principal investigator in the Institute for High-Performance Computing of the Russian Academy of Sciences (Russia) directing research on unifying remote access to high-performance computing systems as a web service and speeding up own neural network modelling software
  • Associated thesis [P67]
[J21] 1994-01 - 1994-06 Research assistant in Moscow Insitute of Physics and Technology (Russia) leading development of simulation and visualization software in a project "Computer simulation of non-linear wave processes in gaseous streams"
[J22] 1994-01 - 1997-06 Research assistant and principal investigator in Moscow Institute of Physics and Technology (Russia) directing research on developing a public research, development and experimentation toolset to design, model and optimize semiconductor neural networks as a practical step towards collaborative reverse engineering of a brain and development of a brain-inspired neuro-computer
  • Project description [M10]
  • Associated publications [P70, P69, P68]
[J23] 1992-02 - 1993-06 Founder and CTO in own research, development and consulting company developing and selling software for the automation of financial reporting in companies

Startups

# Year Description
[C1] 2019-07 - 2020-04 Co-founder and CTO of CodeReef, France (a subsidiary of cKnowledge SAS) building a web-based AI playground.
[C2] 2015-05 - 2019-03 Co-founder, CTO and scientific advisor of dividiti, UK using my CK framework in real life condititions to help companies, universities and non-profit organizations automate their tedious deep tech R&D and accelerate the technology transfer from academia to industry.
[C3] 1992-02 - 1993-06 Founder and CTO of a commercial engineering company developing and selling software for automation of financial reporting in companies (used profits to fund my own research project on brain-inspired computing during undergraduate studies)

Institution building

# Year Description
[I1] 2014-07 - cur. Established the cTuning foundation (non-profit research organization registered in France and supported by companies and the EU FP7/H2020 program and companies) to continue my community-driven development of the open-source cTuning technology (Collective Knowledge infrastructure and repository) as well as the methodology for collaborative and reproducible research and experimentation gradually adopted by the leading ACM, IEEE and NIPS/NeurIPS conferences
  • Developing methodology and tools for artifact evaluation at the leading ACM and IEEE conferences including CGO, PPoPP, PACT, SuperComputing and SysML [O4]
  • Continue supporting my open-source tools and repository for systematic, collaborative and reproducible AI, ML, computer systems and quantum research and experimentation ("big data" predictive analytics, machine learning, data mining, statistical analysis, autotuning, run-time adaptation, sw/hw co-design)
  • Continue validating my new publication model where all research artifacts (tools, benchmarks, data sets, models) are continuously shared, validated and reused by the community [P14,E16,E17,E15,E12,E13]
  • Awards [A3]
  • Guest lectures [L2]
  • Funded by [F3]
  • Associated job [J5]
  • Associated publications [P14]
  • Associated events [E16,E17,E15,E12,E13]
  • Associated software [S2]
  • An associated public repository for collaborative and reproducible experimentation with interactive graphs and articles [R2]
[ Website ]
[I2] 2010-03 - 2011-08 Helped to establish Intel Exascale Lab in France while directing application optimization group
  • On industrial leave from INRIA invited by Intel to help establish new Exascale Lab in France, prepare long-term R&D vision based on my cTuning technology [I3], and lead the team of 8 researchers, engineers, and students
  • Funded by [F5]
  • Associated job [J8]
  • My first team members [Q6, Q4, Q3]
  • Associated publications [P21, P22, P27, P29]
  • Associated events [E30, E31, E32]
  • Associated in-house software and repository [S7, S11]
[I3] 2008-01 - cur. Founded public cTuning.org portal to bring academia and industry together to crowdsource benchmarking, optimization and co-design of computer systems using the public repository of knowledge, plugin-based auto-tuning, run-time adaptation, big data predictive analytics (machine learning, data mining, statistical analysis, feature detection), and collective intelligence
  • Proposed and implemented idea of crowdsourcing program analysis, optimization and run-time adaptation using plugin-based repository and infrastructure for the EU MILEPOST project [F9] that was effectively used to train machine-learning based self-tuning compiler MILEPOST GCC [S10] and meta-optimizer cTuning CC [S9]
  • Initiated new publication model in computer engineering where all research artifacts (tools, benchmarks, datasets, models) are continuously shared, validated and reused by the community [E25]
  • Awards [A4]
  • Keynotes [K2, K1]
  • Funded by [F9, F8, F7, F6, F4]
  • Associated job [J12, J7]
  • Associated publications [P17, P16, P18, P20, P21, P22, P23, P25, P26, P28, P29, P30, P31, P32, P33, P35, P36, P37, P38, P39, P40, P41, P42]
  • Associated events [E23, E25, E25, E29, E36, E38]
  • Associated software [S11, S3, S6, S10, S8, S9, S14, S12, S5, S13, S4]
  • An associated public repository of optimization knowledge [R3]
[ Website ]

Editor

# Year Type Description
[Ed1] 2015-04 Special journal issue Guest editors: Alex Jones (University of Pittsburgh, USA) and Grigori Fursin (INRIA, France).
Special Issue on Reproducible Research Methodologies, IEEE Transactions on Emerging Topics in Computing (TETC)

[ IEEE TETC Website ][ Flyer and CFP ][ Related OCCAM project ][ Related Collective Mind project ][ Related Collective Mind repository ]
[Ed2]


Keynotes

# Year CK Name
[K1] 2013-06 Keynote at iWAPT 2013 / ICCS 2013 in Barcelona, Spain ("Crowdsourcing autotuning: challenges and possible solutions")

[ Website ] [View presentation from CK / cM]
[K2] 2013-03 Keynote at HPCS 2013 in NTU, Taipei, Taiwan ("Systematizing tuning of computer systems using crowdsourcing and statistics")

[ Website ] [Conference flyer from CK / cM] [View presentation 1 from CK / cM] [View presentation 2 from CK / cM]

Social activities and community service

# Year Description
[O1] 2020-12 - cur. A founding member of MLCommons helping to accelerate machine learning innovation

[ MLCommons press-release ]
[O2] 2019-01 - cur. ACM Publications Board's Digital Library committee member

[ ACM DL website ][ New ACM DL twitter ][ ACM policy on Artifact Review and Badging which we co-authored in 2016 based on our experience with Artifact Evaluation at CGO,PPoPP,PACT,SuperComputing and other ACM conferences ]
[O3] 2016-01 - cur. Founding member of the ACM Task Force on Data, Software, and Reproducibility in Publication

[ Website ][ Automation and unification of Artifact Evaluation ][ ACM policy on Artifact Review and Badging which we co-authored in 2016 based on our experience with Artifact Evaluation at CGO,PPoPP,PACT,SuperComputing and other ACM conferences ]
[O4] 2014-11 - cur. Initiator and co-chair of the Artifact Evaluation (validating research results from accepted papers and standardizing Artifact Appendices) for CGO, PPoPP, PACT, SysML and other ACM, IEEE, and NeurIPS (NIPS) conferences
  • Our initiative is supported by ACM since 2015
[ Official website ][ Artifact Appendix (I introduced them to gradually standardize and automate artifact submission and evaluation) ][ Collective Knowledge - I developed it to help researchers make their research artifacts (code, data, models) reusable and automate their experimental workflows ]
[O5] 2009-09 - cur. Supporter of the MSF (Doctors Without Borders)

[ Website ]
[O6] 2008-01 - cur. Founder and CEO of the non-profit cTuning foundation
  • The non-profit cTuning foundation enables and promotes collaborative, reproducible and systematic research, experimentation and development in computer engineering. It develops and supports a public repository of knowledge and related infrastructure. cTuning foundation also helps academic and industrial partners systematize, automate and speed up optimization, benchmarking and co-design of computer systems across all software and hardware layers (applications, compilers, run-time libraries, heterogeneous multi-core architectures) for minimal execution time, power consumption, failures, and other costs while reducing time to market for new solutions. We were among the first researchers to convert this complex task into a unified big data problem and tackle it using open source Collective Mind/Collective Knowledge public repository of knowledge, common plugin-based auto-tuning infrastructure, run-time adaptation, machine learning, data mining, feature selection, crowdsourcing, and collective intelligence
  • I started it in 2008 to continue my R&D from the EU MILEPOST project (2006-2009) [J12] as a community effort
  • Officially registered in France in 2014
[ Website ]

Examiner

  • 2019-06 - PhD examiner for Chandan Reddy (Paris University / ENS Paris)
  • 2015-11 - PhD examiner for Luka Stanisic (INRIA Grenoble, France)
  • 2013-02 - PhD examiner for Ettore Speziale (Politecnico di Milano, Italy)
  • 2013-02 - PhD examiner for Michele Tartara (Politecnico di Milano, Italy)
  • 2013-02 - PhD examiner for Paolo Grassi (Politecnico di Milano, Italy)
  • 2013-02 - PhD examiner for Simone Corbetta (Politecnico di Milano, Italy)

Expert service

# Year Description
[X1] 2019 Invited as an international expert to brainstorm further improvements of the ACM Digital Library[ ACM DL ]
[X2] 2016 Invited as an international expert to prepare and co-author the policy on "Result and Artifact Review and Badging" for the ACM conferences, and extend Artifact Appendices which I originally introduced for CGO and PPoPP[ ACM policy on "Result and Artifact Review and Badging" ][ CGO-PPoPP artifact appendices ][ SC'17 artifact appendix ]
[X3] 2013 Invited as an international expert to review research proposals for the open grant competition of the Russian Federation to attract leading scientists to Russian universities with a total budget of around 200,000,000 euros
[X4] 2013 Invited to contribute to EU HiPEAC roadmap on advanced computing (2013 - 2020) [ Online document ]
[X5] 2012 Invited as an international expert to review research proposals for the open grant competition of the Russian Federation to attract leading scientists to Russian universities with a total budget of around 200,000,000 euros
[X6] 2012 - cur. Consulting several major international IT companies (names are currently under NDA) to design faster and more power efficient production computer systems (software and hardware)
[X7] 2011 Invited as an international expert to prepare common EU-Russia IT call (related to GPGPU, programming models, performance and power tuning)
[X8] 2011 Invited as an international expert to review research proposals for the open grant competition of the Russian Federation to attract leading scientists to Russian universities with a total budget of around 300,000,000 euros
[X9] 2009 Invited to contribute to EU HiPEAC roadmap on advanced computing (2009 - 2020) long-term ideas on collaborative and reproducible computer systems' research based on my cTuning and MILEPOST technology - I continued this initiative as an Artifact Evaluation for the leading conferences including CGO, PPoPP, PACT, RTSS and SC.[ Online document ]
[X10] 2009 Consulting ARC (Synopsys) to apply cTuning and MILEPOST technology for multi-objective tuning (performance/code size/power) of customers' programs
[X11]

Major research achievements

# Year Description
[M1] 2014 - cur. Initiated Collective Knowledge Project aggregating all my past R&D to fix science, enable collaborative and reproducible research, and facilitate transfer of AI, ML and quantum research to industry to solve the world's most challenging problems!
  • ReQuEST: 1st open, reproducible and Pareto efficient SW/HW co-design competition for deep learning (speed, accuracy, costs) uses CK to put ML and systems researchers on the same page and share produced artifacts as reusable, portable and customizable components (open science) [E3]
  • Continued my new publication model in computer engineering [M3] where all research artifacts (tools, benchmarks, datasets, models) are continuously shared and validated by the community, and extended in new cM Lego-style R&D scenarios
  • Awards [A3]
  • Guest lectures [L2]
  • Partially funded by [F3]
  • Associated job [J5]
  • Associated publications [P14, P11, P15, P35, P14]
  • Associated events [E3, E16,E17,E15,E12,E13]
  • Open-source Collective Knowledge Infrastructure [S2]
  • Public Collective Knowledge repository for collaborative and reproducible experimentation with interactive graphs and articles [R2]
  • On-going, community-driven effort
[ Collective Knowledge live repository with interlinked code, data, experimental results, predictive models, interactive graphs and articles, etc. ][ Collective Knowledge Framework for collaborative, reproducible and cost-aware computer engineering ][ Partners / users ][ Use cases ][ ACM project to integrate CK with Digital Library funded by Sloan Foundation ][ Artifact Evaluation ][ Development GitHub Website ][ Documentation ][ Google Group discussions ]
[M2] 2011 - cur. Developed theoretical Collective Mind foundations and supporting plugin-based infrastructure with public web-services to enable collaborative, systematic and reproducible analysis, design and optimization of adaptive computer systems based on extensible public repositories of knowledge, crowdsourcing, online tuning and machine learning, and to initiate new publication model with reproducible results where all research artifacts are continuously shared, validated and extended by the community
  • Aggregates all my past ideas and concepts from [M10, M9, M8, M7, M6, M5, M4]
  • Partially funded by INRIA 4 year fellowship [A4]
  • Continued new publication model in computer engineering [M3] where all research artifacts (tools, benchmarks, datasets, models) are continuously shared and validated by the community, and extended in new cM Lego-style R&D scenarios
  • Publications [P17, P16, P18, P20, P21, P22]
  • Associated events [E25, E23, E27, E30]
  • Collective Mind Framework (plugin-based knowledge management system) [S3]
  • Collective Mind Public repository [R3]
  • Collective Mind framework and repository discontinued for a much smaller and faster Collective Knowledge Infrastructure [M1]
[ Collective Mind live repository ][ Online advice web service to predict optimizations based on features ][ Universal auto-tuning and learning pipeline for top-down multi-objective optimization ]
[M3] 2007 - cur. Evangelized and pushed reproducible research in computer engineering to masses and initiate new publication model with reproducible results where all research artifacts (experimental pipelines, benchmarks, codelets, data sets, tools, models) are continuously shared, validated and extended by the community
  • Since it was extremely difficult to persuade community which mainly focuses on publications to start sharing their research artifacts, in 2007, I developed and opened a cTuning.org web portal with a public repository, and shared all my past experimental results and artifacts to set up an example. Since then this approach was gradually picked up by the community and now being evaluated in major computer science conferences including CGO and PPoPP [E16, E17].
  • Partially funded by [A4,F3]
  • Awards [A3]
  • Publications [P14, P17, P16, P18, P20, P21, P22]
  • Associated events [E16,E17,E15,E12,E13,E25, E23, E27, E30]
  • Open-source Collective Knowledge Infrastructure [S2]
  • Public Collective Knowledge repository for collaborative and reproducible experimentation with interactive graphs and articles [R2]
  • Open-source Collective Mind Framework (plugin-based knowledge management system) [S3]
  • Public Collective Mind Public repository [R3]
  • Our vision paper on community-driven reviewing of papers and artifacts [P14]
  • On-going, community-driven effort backed up by ACM since 2015 ...
[ CrowdTuning portal ][ Collective Mind live repository ][ Online advice web service to predict optimizations based on features ][ Universal auto-tuning and learning pipeline for top-down multi-objective optimization ]
[M4] 2006 - cur. Prepared theoretical foundations and led development of the first practical machine learning based self-tuning compiler (MILEPOST GCC and cTuning CC) and plugin-based multi-objective auto-tuning framework (execution time, code size, compilation time, power consumption or any other user defined metric) combined with collective participation of multiple users (cTuning.org)
  • Aggregates my past ideas and concepts from [M10, M9, M8, M7, M6, M5]
  • Funded by [F9, F6]
  • Collaboration with IBM (Israel), University of Edinburgh (UK), ARC (now Synopsys, UK), CAPS Entreprise (France), and ICT (China)
  • Initiated new publication model in computer engineering [M3] where all research artifacts (tools, benchmarks, datasets, models) are continuously shared and validated by the community within cTuning plugin-based framework [M2]
  • Publications [P53, P52, P47, P46, P42, P41, P37, P36, P33, P31, P26]
  • Associated events [E41, E39, E38, E37, E36, E34, E32]
  • MILEPOST GCC software [S10]; most of technology is now available in mainline GCC and is being added to major commercial compilers
  • cTuning CC software [S9]
  • cTuning repository [R5]
  • Considered by IBM to be the first practical machine-learning based compiler in the world (IBM press-release [P34])
  • All benchmarks, datasets, tools, models and experimental results have been released to public for collaborative validation and extension!
[ cTuning collaborative portal ][ Online advice web service to predict optimizations based on features ]
[M5] 2004 - cur. Developed concept of statistical collaborative optimization (collective optimization) that dramatically speeded up analysis and multi-objective optimization of computer systems by transparently distributing them among multiple users and combining with statistical analysis
[M6] 2004 - cur. Developed concept of split-compilation to statically enable dynamic optimization and adaptation using code multi-versioning and low-overhead run-time adaptation as a reaction to program behavior, architecture changes, and dataset parameters
  • Associated software [S14, S4]
  • Funded by [A8, F6, F7]
  • Supports predictive scheduling for heterogeneous architectures [P39] and adaptive libraries combined with automatically built run-time decision trees [P37]
  • Publications [P57, P48, P45, P38, P35, P32, P28, P17]
  • Highest ranked paper introducing technique [A7, P57]
  • Prepared for mainline GCC during Google Summer of Code program in 2009 [F6]
  • Finalized in mainline GCC 4.8 in 2012
  • Used and extended by Google
[M7] 2004 - cur. Developed concept of making hardwired tools and applications interactive using simple, event-based plugin framework
  • Supports GCC, LLVM, Open64, PathScale compilers; enables self-tuning applications [M6, P39, P17]
  • Funded by [A8, A4, F9, F6, F4]
  • Associated software [S17, S4, S5]
  • Publications [P57, P54, P50, P39, P37, P32, P31, P26, P28, P17]
  • Now available in mainline GCC and extensively used by industry and academia world wide
[M8] 1999 - cur. Developed multiple unconventional interdisiplinary approaches for empirical program and architecture analysis, optimization and co-design through reactions to (possibly semantically non-equivalent) modifications (source or binary instruction, loop or thread level patching)
  • Used to quickly detect application CPU/memory bounds and performance bottlenecks without the need for slow simulators or possibly misleading hardware counters, or characterize programs and arcthiectures similarity through a vector of reactions to canonical transformations to predict most profitable optimizations
  • Funded by [F10, F9, F5]
  • Software [S19, S5, S11, S3]
  • Publications [P63, P62, P59, P58, P46, P38, P28]
  • Used in EU MHAOTEU project [J17], my PhD R&D [Z6], and in all later projects on program and architecture co-design, crowd-tuning and machine learning
  • Adopted and extended in Intel Exascale Lab within DECAN tool (decremental analysis via binary patching) [P29]
[M9] 1997 - 1999 Developed concept of unified access to HPC resources for non-specialists as a simple web service
  • Used to simplify deployment of my parallel (MPI-based) neural network modelling software on distributed and varying HPC resources
  • Partially funded by [A11, J23]
  • Software [S20, S21]
  • Publications [P67, P68, P69, P70]
  • Repository [R8]
  • Used in EU MHAOTEU project [J17], my PhD R&D [Z6] and later work on cTuning.org and crowd-tuning
[M10] 1993 - 1999 Developed prototype of a public research, development and experimentation toolset to design, model and optimize semiconductor neural networks as a practical step towards collaborative reverse engineering of a brain and development of a brain-inspired neuro-computer
  • Partially funded by [A11, J23]
  • Hardware: [H1] (own ADC/DAC board for automatic measurement of characteristics of semiconductor (neural) devices)
  • Software [S21]
  • Publications [P68, P69, P70]
  • Repository [R8]
  • Since modeling was already too slow, unreliable and power hugry, it forced me to switch most of the following R&D effort to make program and arcthiecture optimization practical and combine it with statistical analysis and machine learning using interdisiplinary background
  • Ideas from this project was reused in all my next interdisiplinary projects on program and architecture characterization, optimization and co-design using machine learning and crowdsourcing

Public or in-house repositories of knowledge

# Year CK Description
[R1] 2017-01 - cur. Open repository of reusable, customizable, portable and optimized AI artifacts to accelerate AI research and boost innovation
  • Developed by the non-profit cTuning foundation and dividiti
  • Opened to public in 2017
  • Supports our 1st open, reproducible and Pareto efficient SW/HW co-design competition for deep learning (speed, accuracy, costs) at ACM ASPLOS [E3]
  • Powered by Collective Knowledge Framework [S2]
[ cKnowledge.org/repo ][ Portable, customizable and reusable AI CK components ][ Locator of CK-based AI artifacts (beta) ][ CK repositories ]
[R2] 2014-11 - cur. Collective Knowledge public repository (CK aka cTuning4) to continue improving whole experimental setup sharing (code, data, dependencies, experimental results, models) along with interactive articles
  • Developed by the non-profit cTuning foundation
  • Opened to public in 2015
  • Received award [A3]
  • Included all past and current semantically connected research artifacts from my research, development and experimentation (hundreds of codelets and benchmarks; thousands of datasets; GCC, LLVM, Open64, PathScale, ICC compiler optimization description; tools and scripts; online tuning plugins; machine learning plugins; adaptive exploration pluigns; graph plotting plugins; data mining plugins; machine learning based meta compiler; MILEPOST GCC, etc
  • Supports our initiatives on Artifact Evaluation and new publication models where results and papers are validated and improved by the community [M3]
  • Partially funded by [F3]
  • Powered by Collective Knowledge Framework [S2]
[ Collective Knowledge live repository ][ Examples of interactive graphs ][ Examples of interactive reports ]
[R3] 2011-09 - cur. Collective Mind public repository (cM aka cTuning3) to start collaborative systematization of analysis, design and optimization of computer systems based on extensible public repositories of knowledge, crowdsourcing, online tuning and machine learning, and to initiate new publication model where all research artifacts are continuously shared, validated and extended by the community
  • Opened to public in 2013
  • Included all past and current semantically connected research artifacts from my research, development and experimentation (hundreds of codelets and benchmarks; thousands of datasets; GCC, LLVM, Open64, PathScale, ICC compiler optimization description; tools and scripts; online tuning plugins; machine learning plugins; adaptive exploration pluigns; graph plotting plugins; data mining plugins; machine learning based meta compiler; MILEPOST GCC, etc
  • Connected with Android Collective Mind Node [S6] to crowdsource program and architecture characterization and multi-objective autotuning (execution time, code size, compilation time, power consumption) using any available Android-based mobile phone, tablet or laptop
  • Used for the new publication model [E25]
  • Funded by [A4, F4]
  • Powered by Collective Mind Framework [S3]
[ Collective Mind live repository ][ Online advice web service to predict optimizations based on features ][ Universal autotuning and learning pipeline for top-down multi-objective optimization ]
[R4] 2010-03 - 2011-08 In-house Codelet Tuning Repository for Intel Exascale Lab (aka cTuning2) to decompose large applications into codelets for continuous characterization and tuning
  • Developed with my team [Q6, Q4, Q3] as customizable repository for Intel Exascale Lab, CEA, GENCI and UVSQ (France)
  • Funded by [F5]
  • Powered by Codelet Tuning Infrastructure [S7]
  • Discontinued for [R3]
[R5] 2006-01 - cur. cTuning.org public repository (aka cTuning1) to start collaborative systematization of analysis, design and optimization of computer systems based on extensible public repositories of knowledge, crowdsourcing, online tuning and machine learning
  • Opened to public in 2008
  • Included past research experimentatal results on program and architecture multi-objective tuning (execution time, code size, compilation time, power consumption) for reproducibility and collaborative extensions
  • Connected with MILEPOST GCC [S10] for continuous and online training and improvement of the prediction models
  • Funded by [F9, F8, F7, F6, F4]
  • Powered by cTuning Framework [S11]
  • Gradually being discontinued for [R3]
[ cTuning live repository ][ Online advice web service to predict optimizations based on features ]
[R6] 2004-06 - 2006-06 In-house collaborative optimization repository for research on multi-objective program and architecture autotuning and co-design combined with machine learning
  • Powered by FCO framework [S15]
  • Funded by [A8, F8]
  • Discontinued for [R5]
[R7] 1999-02 - 2006-06 In-house collaborative optimization repository for research on multi-objective program and architecture characterization, optimization and co-design with first experiments on predictive modeling
  • Powered by EOS framework [S19]
  • Developed and used in the EU FP5 MHAOTEU project [J17]
  • Funded by [F10, A8]
  • Discontinued for [R6]
[R8] 1993-03 - 1999-02 In-house Experimental Repository for research, development and experimentation on novel, semiconductor neural networks, and on providing unified access to HPC resources as a web service
  • Powered by SCS framework [S20]
  • Partially funded by [A11, J23]
  • Discontinued for [R7]

Awards, prizes and fellowships

# Year Description
[A1] 2017-02 Test of time award for our CGO'07 paper on "rapidly selecting good compiler optimizations using performance counters"
  • Based on the positive feedback from the community, I continued this research as a community effort by developing a supporting open-source technology and a public repository to crowdsource machine-learning based optimization of realistic workloads across diverse hardware and data sets (Collective Knowledge).
  • This framework now assists various Artifact Evaluation initiatives at the premier ACM conferences on parallel programming, architecture and code generation (CGO, PPoPP, PACT, SC), which aim to encourage sharing of code and data to enable collaborative and reproducible systems research
  • Dividiti Ltd is also commercializing this technology with ARM, GM and other companies.
[ Blog article about our award ][ Collective Knowledge framework for customizable, multi-objective and ML-based tuning ][ Public Collective Knowledge repository to crowsource optimization ][ Collaborative AI optimization ][ CK-powered projects in dividiti Ltd ][ CK-powered performance analysis and optimization at ARM ][ CK-powered Caffe optimization with GM ][ Artifact evaluation for CGO,PPoPP,PACT,RTSS and SC ][ Wikipedia article about CK ]
[A2] 2016-05 Microsoft Azure Research award for Collective Knowledge Technology (crowdsourcing application benchmarking and optimization via Azure cloud)
[A3] 2014-12 EU HiPEAC technology transfer award for transferring cTuning technology to ARM to systematize benchmarking and combining it with "big data" predictive analytics
[A4] 2012-04 - 2014-10 INRIA award and fellowship for "making an outstanding contribution to research" for making program and architecture optimization and co-design more practical, systematic and reproducible by combining cTuning plugin-based autotuning technology with statistical analysis, machine learning and community-driven curation
[A5] 2010-11 HiPEAC award for paper "Evaluating Iterative Optimization across 1000 Data Sets" [P30], PLDI 2010, Canada
[A6] 2009-12 HiPEAC award for paper "Portable Compiler Optimization Across Embedded Programs and Microarchitectures using Machine Learning" [P33], MICRO 2009, NY, USA
[A7] 2005-01 Highest ranked paper "A practical method for quickly evaluating program optimizations" [P57] at HiPEAC 2005, Barcelona, Spain
[A8] 2004-10 - 2005-11 EU HiPEAC Fellowship to collaborate with INRIA Saclay (France)
[A9] 2000-01 - 2001-12 International Overseas Research Student Award (fellowship) for PhD research from the UK government
[A10] 1999-06 Golden Medal for MS studies and thesis from Moscow Institute of Physics and Technology (Russia)
[A11] 1996-08 - 1998-08 International Soros Science Education Program fellowship "in recognition and appreciation of outstanding achievements in the study of science at the university level" (~300 students in Russia)

Major funding

# Year Description
[F1] 2016-05 - cur. Microsoft Azure Research Sponsorship for the cTuning foundation to move our Collective Knowledge Repository (cknowledge.org/repo) to Microsoft cloud

[ Collective Knowledge Repository ]
[F2] 2016-05 - 2016-07 EU FP7 609491 TETRACOM funding for Imperial College London and dividiti to crowdsource OpenGL testing and bug detection using open-source Collective Knowledge Framework and repository

[ open-source, BSD-licensed, Collective Knowledge Technology ][ public CK repository ][ ICL multicore programming group ][ dividiti ]
[F3] 2014-11 - 2015-04 EU FP7 609491 TETRACOM funding to validate open-source cTuning technology in ARM (open-source infrastructure and repository for collaborative and reproducible autotuning and "big data" predictive analytics)

[ Developed Collective Knowledge Infrastructure and Repository (by non-profit cTuning foundation) ]
[F4] 2013-07 - 2013-10 HiPEAC industrial internship funding for Abdul Memon to validate PhD results in STMicroelectronics (France) working on "Auto Tuning Optimization System Acceleration for Embedded Linux Stacks"

[ HiPEAC description ]
[F5] 2010-04 - 2011-08 Funding from Intel and CEA to invited to help establish new Intel Exascale Lab in France based on cTuning technology, serve as a head of application characterization and optimization group, and direct research and development
[F6] 2009-06 - 2009-08 Funding from Google (GSOC program) to move cTuning and MILEPOST technology to mainline GCC (Interactive Compilation Interface and code multiversioning to make statically compiled programs adaptable at run-time) for 2 students from ICT (China)

[ GSOC description page for adding and extending plugin-based framework (ICI) to mainline GCC ][ GSOC description page for statically enabling dynamic optimizations in GCC through code cloning ]
[F7] 2008-03 - 2008-05 EU HiPEAC funding for Victor Jimenez (PhD student from UPC, Spain) to visit my research group in INRIA Saclay (France)
  • Extending my previous work on statically enabling dynamic optimization [P57] with predictive scheduling for heterogeneous architectures [P39]
[F8] 2008-01 - 2008-04 Funding from ICT (China) for students and faculty exchange to extend cTuning and MILEPOST technology
[F9] 2006-09 - 2009-09 Funding from EU FP6 MILEPOST project to develop practical machine learning based self-tuning compiler
  • I was one of the initiators of this project responsible for INRIA part - development of a repository of knowledge, common plugin-based auto-tuning infrastructure, statistical analysis, predictive modeling, GCC plugin-based framework
[F10] 1999-02 - 2001-12 Funding from EU FP5 MHAOTEU project to develop tools for memory hierarchy optimization for High-Performance Computer Systems
  • I was responsible for University of Edinburgh part developing polyhedral source-to-source compiler, plugin-based distributed analysis and auto-tuning infrastructure and a repository of knowledge while reusing some of my M.S. developments for unifying access to supercomputers through the web
[F11]


Education

# Year Degree/course
[Z1] 2020-11 Machine Learning by Andrew Ng (Stanford University)

[ Certificate ]
[Z2] 2013-10 Participating in Dagstuhl seminar on "Automatic Application Tuning for HPC Architectures" in Germany

[ Website ]
[Z3] 2008-07 Attending 4th International Summer School on Advanced Computer Architecture and Compilation for Embedded Systems (ACACES 2008) organized by HiPEAC in L'Aquila, Italy
  • Associated poster [P41]
[ Website ][ Courses ]
[Z4] 2007-07 Attending 3rd International Summer School on Advanced Computer Architecture and Compilation for Embedded Systems (ACACES 2007) organized by HiPEAC in L'Aquila, Italy
  • Associated poster [P44]
[ Website ][ Courses ]
[Z5] 2006-07 Attending 2nd International Summer School on Advanced Computer Architecture and Compilation for Embedded Systems (ACACES 2006) organized by HiPEAC in L'Aquila, Italy

[ Website ][ Courses ]
[Z6] 2004-05 PhD in computer science from the University of Edinburgh, UK
  • Awards: ORS award
  • Advisor: Prof. Michael O'Boyle (University of Edinburgh, UK)
  • Examination board: Francois Bodin (CAPS Entreprise, France), Marcelo Cintra (University of Edinburgh, UK)
  • Associated thesis [P58]
[Z7] 1999-06 MS summa cum laude in computer engineering from Moscow Institute of Physics and Technology and Institute of High-Performance Computing Systems of Russian Academy of Sciences, Russia
  • Awards: golden medal
  • GPA=4.0/4.0; TOEFL=593; GRE=x/800/780
  • Associated thesis [P67]
[Z8] 1997-06 BS summa cum laude in physics and electronics from Moscow Institute of Physics and Technology, Russia
  • Interdisciplinary courses in physics, electronics, mathematics, statistics, machine learning, brain simulation and computer science
  • Associated publications [P70, P69, P68]
[Z9] 1993-06 Graduated from Moscow Secondary School No249 with medal

[Z10] 1993-06 Graduated from Moscow College of Physics and Technology, Russia
  • GPA=4.0/4.0

Major software and datasets
# Year CK Description
[S1] 2015-03 - cur. Android application "Crowdsource Experiments" to for collaborative hardware and software optimization and machine learning using Android-based mobile devices provided by volunteers
  • Released in 2016
  • Connected to public Collective Knowledge Repository [R2]
  • Associated publications [P13, P11]
  • 1st open, reproducible and Pareto efficient SW/HW co-design competition for deep learning (speed, accuracy, costs) uses CK to put ML and systems researchers on the same page and share produced artifacts as reusable, portable and customizable components (open science) [E3]
[ Program at Google Play Website ][ Related program for deep learning at Google Play Website ]
[S2] 2014-11 - cur. Collective Knowledge Framework and Repository (CK aka cTuning4, BSD-license) - Collective Knowledge (CK) is a light-weight, portable, modular and python-based framework, repository, web service and SDK to organize, describe, cross-link and share user code, data, experimental setups and meta information as unified and reusable components with JSON API via standard Git services (such as GITHUB or BitBucket).
  • Opened to public in 2015
  • Pre-released in May, 2015 (V1.2, permissive and simplified BSD license)
  • Partially funded by EU FP7 TETRACOM 6-months grant [A4]
  • Supports our new publication model in computer engineering where all research artifacts (tools, benchmarks, datasets, models) are continuously shared and validated by the community [P14,E16,E17,E15,E12,E13]
  • Awards [A3]
  • Guest lectures [L2]
  • Associated publications [P14, P11, P15, P35, P14]
  • Associated events [E16,E17,E15,E12,E13]
  • Associated live repository [R2] - hosting in Azure is sponsored by Microsoft
  • ReQuEST: 1st open, reproducible and Pareto efficient SW/HW co-design competition for deep learning (speed, accuracy, costs) uses CK to put ML and systems researchers on the same page and share produced artifacts as reusable, portable and customizable components (open science) [E3]
[ Website ][ Partners / users ][ Use cases ][ Public repository to crowdsource optimizations ][ ACM project to integrate CK with Digital Library funded by Sloan Foundation ][ Artifact Evaluation ][ Development GitHub Website ][ Documentation ][ Google Group discussions ]
[S3] 2011-09 - 2013-09 Collective Mind Framework and Repository (cM aka cTuning3) - plugin-based knowledge management system to preserve, systematize and share all research, development and experimentation artifacts using private or in-house web and plugin-based, customizable, schema-free, NoSQL repository of knowledge combined with crowdsourcing and machine learning; collaborative and agile implementation and systematization of experimental scenarios combined with statistical analysis and data mining; plugin-based program and architecture autotuning and co-design combined with crowdsourcing, machine learning and run-time adaptation
  • Released in 2013 (V1.0beta, standard BSD license)
  • Partially funded by INRIA 4 year fellowship [A4]
  • Includes software [S4, S5, S6]
  • Aggregated and unified all my past research and development ideas and prototypes within new Collective Mind Framework and Repository to systematize collaborative research, development and experimentation
  • Continued new publication model in computer engineering where all research artifacts (tools, benchmarks, datasets, models) are continuously shared and validated by the community, and extended in new cM Lego-style R&D scenarios
  • Technology: easily customizable for any end-user R&D scenario through plugins; agile development methodology; NoSQL databases with JSON representation; ElasticSearch indexing; unified python plugins for web-services, autotuning, data mining and machine learning; OpenME interface to connect applications and tools written in C, C++, Fortran, PHP to cM; upport for practically any Unix and Windows-based desktops, laptops, supercomputers, cloud servers, and even tablets and mobile phones with ARM,Intel,ST,Loongson,AMD,NVidia and other chips; powerful graph capabilities
  • Publications [P17, P16, P18, P20, P21, P22]
  • Associated events [E25, E23, E27, E30]
  • Associated live repository and online advice service [R3]
[ Website ][ cM SVN download ][ cM wiki ][ New cM development tasks ][ Old cM development tasks ][ cM Google Group discussions ]
[S4] 2011-09 - cur. OpenME event-based plugin framework and unified interface to making rigid, hardwired applications and tools interactive, tunable and adaptive
  • Released in 2013
  • Technology: C, C++, Fortran, PHP, python event-based plugin framework
  • Included plugins to open up latest GCC and LLVM for fine-grain program analysis and autotuning, and to open up HPC applications for online tuning and adaptation on heterogeneous architectures [P17, P39]
  • Current version: 1.0beta (LGPL v2 license)
  • Associated publications [P17, P16, P18, P20, P21, P22]
[ Website ]
[S5] 2011-09 - cur. Alchemist plugin for fine-grain program feature extraction, decremental analysis, and optimization in GCC
  • Collaboration with STMicroelectronics [Q7, F4]
  • Pre-released in 2013
  • Current version: 1.0beta (GPL v2 license)
  • Technology: GCC dynamic plugin written in C
  • Associated publications [P17, P16, P18, P20, P21, P22]
[ Website ]
[S6] 2011-09 - cur. Android Collective Mind Node to crowdsource program and architecture characterization and multi-objective autotuning (execution time, code size, compilation time, power consumption) using any available Android-based mobile phone, tablet or laptop
  • Released in 2013
  • Connected to public Collective Mind Repository [R3]
  • Associated publications [P17, P28, P35, P37]
[ Google Play Website ]
[S7] 2010-03 - 2011-08 Colaborative Codelet Tuning Infrastructure (CTI aka cTuning2) to decompose large applications into codelets for continuous characterization and tuning
  • Collaboration with Intel, CEA, GENCI and UVSQ
  • Developed the concept based on my cTuning technology and developed first prototype with my team [Q6, Q4, Q3] as Intel Exascale Lab in-house autotuning infrastructure and repository extending cTuning1 framework and repository
  • Funded by [F5]
  • Associated job [J8]
  • More info about this activity [I2]
  • Technology: mixed MySQL and NoSQL database; customizable python, PHP and C plugin and web-based infrastructure
  • Availability: private use - after 2012 some parts developed by my team members became available under GPL v3 license
  • Discontinued for [S3]
[S8] 2010-01 - 2011-09 KDatasets to Multiple datasets for cBench [S12] (1000 per benchmark)
  • Released in 2010
  • Last version: 1.0 (GPL v2 license)
  • Funded by [F8]
  • Partially merged with Collective Mind repository [R3]
[ Website ]
[S9] 2009-01 - 2010-03 cTuning CC - machine learning based self-tuning meta-optimizer for any compiler including GCC, LLVM, ICC, etc.
  • Released in 2010
  • Last version: 2.5 (GPL v2 license)
  • Moved to Collective Mind Framework as a ctuning.compiler plugin [S3]
  • IBM press-release [P34]
[ Website ]
[S10] 2006-07 - 2009-06 MILEPOST GCC - machine learning based self-tuning compiler
  • Released in 2009
  • Last version: 2.5 (GPL v2 license)
  • Merged with cTuning CC [S9]
  • Funded by [F9]
  • Collaboration with IBM (Israel), University of Edinburgh (UK), ARC (now Synopsys, UK), CAPS Entreprise (France), and ICT (China)
  • Developed within EU FP6 MILEPOST project [J12]
  • More info about this activity [I3]
  • Associated live repository and online advice service [R5]
  • Considered by IBM to be the first practical machine-learning based compiler in the world (IBM press-release [P34])
[ Website ]
[S11] 2006-01 - 2010-03 cTuning and CCC (Continuous Collective Compilation) frameworks for collaborative user-defined program and architecture characterization, multi-objective optimization (execution time, code size, compilation time, power consumption) and co-design of computer systems using public repositories of knowledge, plugin-based autotuning, machine learning and crowdsourcing
  • Released in 2008
  • Current version: 2.5 (GPL v2 license)
  • Collaboration with IBM (Israel), University of Edinburgh (UK), ARC (now Synopsys, UK), CAPS Entreprise (France), and ICT (China)
  • Developed within EU FP6 MILEPOST project [J12]
  • More info about this activity [I3]
  • Technology: mixed MySQL and NoSQL database; customizable PHP,C,C++ plugin and web-based infrastructure
  • Discontinued for [S3]
  • IBM press-release [P34]
[ Development website ][ Tools ]
[S12] 2006-01 - 2010-03 cBench (Collective Benchmark) to unify and systematize benchmarking with multiple datasets for realistic and collaborative program and architecture autotuning and co-design combined with machine learning; unify training and tuning of MILEPOST GCC []; enable research on split compilation []
  • Released in 2008
  • Last version: 1.1 (GPL v2 license)
  • Discontinued and merged with Collective Mind repository [R3]
[ Website ]
[S13] 2006-01 - 2010-03 cDatasets (Collective Datasets) to multiple datasets for cBench [S12] (20..100 per benchmark)
  • Released in 2008
  • Last version: 1.1 (GPL v2 license)
  • Discontinued and merged with Collective Mind repository [R3]
[ Website ]
[S14] 2004-10 - 2010-03 UNIDAPT - universal plugin-based program run-time adaptation framework
  • Software support for my technique for split compilation (to dynamic optimization and adaptation for statically compiled programs using multi-versioning and light run-time adaptation mechanisms) [P57, P45]
  • Supports predictive scheduling for heterogeneous architectures [P39] and adaptive libraries combined with automatically built run-time decision trees [P37]
  • Funded by [A8, F6]
  • Publications [P57, P39, P37, P48, P45, P38, P35, P32, P28, P17]
  • Finalized in mainline GCC 4.8 in 2012
  • Prepared for mainline GCC during Google Summer of Code program in 2009 [F6]
  • Redesigned and unified in [S4]
[ Website ][ GCC multi-versioning description ]
[S15] 2004-06 - 2006-06 Framework for Continuous Optimization (FCO)
  • Released in 2006
  • Licence: GPL v2
  • Collaboration with Institute of Computing Technology (China) to tune applications and compilers for LoongSon and Godson processors [F8, J9]
  • Discontinued for S11
[ Old description ][ Framework with ICI for Open64 compiler ][ Framework with ICI for PathScale compiler ]
[S16] 2004-06 - 2004-12 Code, data and experiment sharing tool to decentralize collection of huge amount of experiments during tuning GCC optimization heuristic
  • I started implementing a plugin framework and a plugin for GCC to transparently tune optimization heuristic and embed special function to collect run-time info about behavior of real programs and data sets in real environments from multiple users. Originally, I connected my framework to MySQL database, but it could not cope with huge amount of data. Therefore, I tried to implement another solution - sharing best optimizations/speedups via P2P networks such as overnet, bittorent and edonkey. Unfortunately, the solution became rather complex and unstable due to a lack of stable and universally acceptable P2P tools. So, I had to move back to MySQL database in the EU FP6 MILEPOST project [S10]. However, I later implemented my own P2P sharing mechanism with noSQL Hadoop-based repository in [S2, S3] while investigating capabilities of new third-party P2P tools.
[S17] 2004-06 - 2009-06 Interactive Compilation Interface (ICI) to open up production compilers (GCC, Open64, PathScale, etc.) through light-weight event-based plugin framework and transform them into powerful interactive research toolsets
  • Publications [P57, P54, P50, P39, P37, P32, P31, P26, P28, P17]
  • Released in 2006; Merged with mainline GCC in 2009 [F6]
  • Last version: 2.5 (GPL v2 license)
  • Redesigned and unified in [S4]
[ Website ]
[S18] 1999-02 - 2006-06 Source-to-source polyhedral transformation server - source-to-source C and Fortran polyhedral transformation server based on MARS compiler
  • First released in 2001
  • Developed during EU FP5 MHAOTEU project [F6] and PhD studies [Z6]
  • Last version: V1.15i (GPL v2 license)
  • Used in first experiments to predict complex optimizations (not just 1 transformation) using machine learning and program semantic and dynamic features (hardware counters)
  • Discontinued for compilers with Interactive Compilation Interface or OpenME [S15, S17, S4]
[ Description ] [Source-to-source transformation server V1.15i for SPARC (~3.9Mb) from CK / cM] [Source-to-source transformation server V1.15i for x86 (~1.2Mb) from CK / cM] [EOS hill-climbing autotuning example (padding, tiling, unrolling) (jpg) from CK / cM]
[S19] 1999-02 - 2006-06 Edinburgh Optimizing Software (EOS) - plugin-based client-server program and architecture characterization and autotuning framework
  • Released in 2001
  • Uses source-to-source C and Fortran polyhedral transformation server based on MARS compiler [S18]
  • Developed during EU FP5 MHAOTEU project [F6] and PhD studies [Z6]
  • Last version: V2.2 (GPL v2 license)
  • Technology: NoSQL based database; java and C plugins; socket communication between modules; java based GUI
  • Included plugins for program memory/CPU characterization through semantically non-equivalent assembler/binary patching [P63, P62, P59, P58, P29]; own source-to-source compiler; fine-grain autotuning plugins (unrolling, array padding) with partial polyhedral optimization support (tiling, fusion/fission, vectorization)
  • Used in first experiments to predict complex optimizations (not just 1 transformation) using machine learning and program semantic and dynamic features (hardware counters)
  • Discontinued for S15 and later for S11
[ Description ] [Download tar.gz package (~6Mb) from CK / cM] [Screenshot (gif) from CK / cM] [EOS architecture (gif) from CK / cM] [EOS hill-climbing autotuning example (padding, tiling, unrolling) (jpg) from CK / cM]
[S20] 1997-06 - 1999-02 SuperComputer Service (SCS) - framework to provide and unify remote access to high-performance computing systems for non-specialists as a simple web service
  • Released in 1999
  • Last version: V1.3 (GPL license)
  • Technology: MySQL based database; java, perl, C, Visual Basic, Visual C modules; standard http/ftp communication; web-based GUI
  • Used to simplify execution of my neural network modelling software on distributed and varying HPC resources
  • Publications [P67, P68, P69, P70]
  • Repository [R8]
  • Partially funded by [A11, J23]
  • Discontinued for S19
[Download tar.bz2 package (~10Mb) from CK / cM] [Shapshot 1 (png) from CK / cM] [Shapshot 2 (png) from CK / cM] [Shapshot 3 (png) from CK / cM] [Shapshot 4 (png) from CK / cM]
[S21] 1993-02 - 1999-02 Semiconductor Brain - semiconductor and modelled neural networks with my own ADC/DAC PC board and analysis software
  • Released in 1997
  • Last version: V2.1 (GPL license)
  • Works with special hardware (ADC/DAC board): [H1]
  • Validated by improving recognition and restoration of characters by neural network in noised environments
  • Technology: Visual Basic and assembler; PSpise analog circuit and digital logic simulation software; MPI for HPC
  • Partially funded by [A11, J23]
  • Publications [P68, P69, P70]
  • Repository [R8]
  • Due to very slow simulation and limiation of semiconducor technology decided to switch to program and architecture optimization to eventually enable fast and low-power neural networks and machine learning
[Download seminconductor neural elements measurement software (ECT) for DOS (~200Kb) from CK / cM] [Download seminconductor neural elements measurement software (ECT) for Windows (~1.6Mb) from CK / cM] [Download sources of the seminconductor neural elements measurement software (ECT) for Windows (~300Kb) from CK / cM] [Shapshot 1 (jpg) from CK / cM] [Shapshot 2 (jpg) from CK / cM]
[S22] 1991-08 - 1991-08 Productivity tool to automatically pack files to remote and removable disks with limited space using fast and random strategy instead of greedy one
  • My first autotuning experience - at that time, we had removable disks with only 360Kb, and it was extremely challenging to archive files on multiple disks. As a practical solution, I created a tool that checked free space on all disks and size of all files to randomly find the best packing strategy (within amount of time specified by a user). Interestingly, it worked as good as a greedy algorithm but about an order of magnitude faster which was important on very slow computers of that time. This was my first successful autotuning experience that I reused in all my further research.

Hardware

# Year CK Description
[H1] 1993-02 - 1999-02 ADC/DAC board for personal computer - automating measurement of charateristics of semiconductor devices
  • Released to MIPT colleagues in 1997
  • Related DOS/Windows software: [S21]
  • Used in various undergraduate MIPT laboratories
  • Partially funded by [A11, J23]
  • Publications [P68, P69, P70]

Talks

Participating in conference and program committees; reviewing

I now manage an international non-profit cTuning foundation and my dividiti startup, and have very little time for reviewing. Nevertheless, I lead a community effort to develop an open methodology and supporting technology for collaborative and reproducible research as a part of Artifact Evaluation at the leading ACM and IEEE conferences including CGO,PPoPP,PACT,SuperComputing.
  • CGO 2019 (PC and reviewer)
  • ASPLOS 2018 (organizing committee)
  • AE CGO-PPoPP 2017 (open reviewing)
  • DATE 2017 (reviewer)
  • DATE 2016 (reviewer)
  • CCGrid 2015 (PC and reviewer)
  • IEEE TPDS journal 2015 (reviewer)
  • ACM TACO journal 2015 (reviewer)
  • ACM TACO journal 2014 (reviewer)
  • PARCO journal 2014 (reviewer)
  • CGO 2014 (external reviewer)
  • IPDPS 2014 (PC and reviewer)
  • IEEE MICRO 2013 (reviewer)
  • SBAC-PAD 2013 (PC and reviewer)
  • MuCoCoS 2013 (PC and reviewer)
  • PACT 2013 (reviewer)
  • CC 2013 (PC and reviewer)
  • IPDPS 2013 (PC and reviewer)
  • MuCoCoS 2012 (PC and reviewer)
  • PhD Thesis, Michele Tartara, Politecnico di Milano, Italy, 2012 (reviewer)
  • HiPEAC/ACM TACO 2012 (PC and reviewer)
  • MASCOTS 2011 (reviewer)
  • CPE 2011 (reviewer)
  • ACM OSR (reviewer)
  • CGO 2011 (PC and reviewer)
  • GROW 2011 (PC and reviewer)
  • IEEE TEVC 2010 (reviewer)
  • Springer-Verlag book on auto-tuning techniques 2010 (reviewer)
  • CLSS 2009 (reviewer)
  • CGO 2010 (reviewer)
  • TPDS 2009 (reviewer)
  • GROW 2010 (PC and reviewer)
  • ICPADS 2009 (PC and reviewer)
  • iWAPT 2009 (PC and reviewer)
  • Open64 2009 (PC and reviewer)
  • JPDC 2009 (reviewer)
  • GROW 2009 (reviewer)
  • JPDC 2008 (reviewer)
  • CGO 2008 (reviewer)
  • IPDPS 2008 (PC and reviewer)
  • IJHPSA 2008 (reviewer)
  • DATES 2008 (reviewer)
  • HiPEAC 2007 (reviewer)
  • ACM ICS 2007 (reviewer)
  • CASES 2007 (PC and reviewer)
  • Transactions on HiPEAC 2006 (reviewer)
  • DATES 2006 (reviewer)
  • PACT 2005 (reviewer)

Teaching and organizing courses

In my spare time, I am preparing new online lectures combined with my Collective Knowledge infrastucture and live optimization repository for open, collaborative, reproducible, automatic and multi-objective software/hardware co-design for emerging workloads including deep learing and AI. I also co-organize the 1st open, reproducible and Pareto-efficient AI/SW/HW co-design tournament (see ReQuEST'18). I also occasionally give guest lectures so if your organization is interested, please get in touch!

# Year CK Type Name
[L1] 2015-11 Lecturer Guest lecture on "Collective Knowledge Technology: from ad hoc computer engineering to collaborative and reproducible data science", University of Manchester (UK)

[ Online info ][ Slides ]
[L2] 2015-03 Lecturer Guest lecture on "systematic, collaborative and reproducible experimentation in computer engineering via Collective Knowledge Framework and Repository", University of Copenhagen (Denmark)
[L3] 2013-03 Lecturer Guest lectures on "systematizing tuning of computer systems using crowdsourcing and statistics", National Taiwan University (Taipei, Taiwan)

[ Lecture 1 sides ][ Lecture 2 slides ]
[L4] 2008-09 - 2008-11 Lecturer Organizing and teaching MS course on "Future computing systems", University Paris Sud (France)

[Course overview from CK / cM] [Lecture 1 (Iterative compilation) from CK / cM] [Lecture 2 (Online tuning and machine learning) from CK / cM] [Lecture 3 (ATF vs ISS) from CK / cM] [Lecture 4 (Dependencies) from CK / cM] [Lecture 5 (Tiling) from CK / cM] [Lecture 6 (Fine grained parallelization) from CK / cM] [Lecture 7 (Coarse grained parallelization) from CK / cM]
[L5] 2007-10 Lecturer Preparing and teaching MS course on "Adaptive and Feedback Driven Compilation and Optimization; Machine Learning", University Paris Sud (France)

[Online lecture from CK / cM]
[L6] 2005-10 Lecturer Preparing and teaching MS course on "Adaptive and Feedback Driven Compilation and Optimization; Machine Learning", University Paris Sud (France)

[Online lecture from CK / cM]

Advising/collaborating
# Year Type Name
[Q1] 2013-07 - 2013-09 Intern Vincent Grevendonk in ARM (UK)
[Q2] 2012-09 - 2013-05 MS student Michael Pankov in Bauman Moscow State Technical University (France)
  • Discussing model-driven optimization search space exploration for compilers
[Q3] 2011-01 - 2011-08 Postdoctoral researcher Pablo Oliveira in Intel/CEA Exascale Lab (France)
  • Collaborating on program characterization and adaptive auto-tuning combined with machine learning extending cTuning and MILEPOST technology
  • Most of this work was under NDA until 2012
  • Presentation of our activities [E30]
[Q4] 2010-04 - 2011-08 MS student and expert engineer Frank Talbart in Intel/CEA Exascale Lab (France)
  • Helping to develop Colaborative Codelet Tuning Infrastructure (cTuning2 aka CTI)
  • Most of this work was under NDA until 2012
  • Presentation of our activities [E30]
[Q5] 2010-04 - 2011-08 PhD student Souad Koliai in UVSQ (France) funded by Intel/CEA Exascale Lab (France)
  • Collaborating to extend my program behavior characterization technique via semantically non-equivalent binary patching [P58, P59, P62, P63] within new DECAN framework
  • Most of this work was under NDA until 2012
  • Associated publication [P29]
[Q6] 2009-09 - 2013-07 PhD student Yuriy Kashikov in UVSQ (France) funded by INRIA and Intel/CEA Exascale Lab (France)
  • Thesis "A holistic approach to predict effective compiler optimizations using machine learning" extending cTuning and MILEPOST technology
  • Prof. William Jalby kindly agreed to be Yuriy's co-advisor while I was preparing HDR in France
  • Associated publications [P24, P26, P27, P32]
  • Associated events [E28, E31]
[Q7] 2009-09 - 2016-06 PhD student Abdul Wahid Memon in the University of Paris-Saclay (France) funded by national government, INRIA and HiPEAC/STMicroelectronics internship
  • Graduation: 17 June 2016
  • Thesis 'Crowdtuning : Towards Practical and Reproducible Auto-tuning via Crowdsourcing and Predictive Analytics' is extending cTuning and MILEPOST technology
  • Associated publications [P14, P16, P26]
  • Related Collective Mind infrastructure and repository (deprecated for CK) [S3]
  • Most of the software, data sets and experiments are shared in the Collective Mind repository to ensure reproducibility and extension of this work by the community
  • research and development continued in the new Collective Knowledge Framework developed by the cTuning foundation (France) and dividiti (UK) [S2]
[Q8] 2009-06 - 2009-08 MS student Yuanjie Huang in Institute of Computing Technology (China) funded by Google Summer of Code
  • Extending MILEPOST technology within GCC to enable fine-grain program optimization
  • Associated publications [P23, P30, P32]
[ Google Summer of Code description page ]
[Q9] 2009-06 - 2009-08 MS student Liang Peng in Institute of Computing Technology (China) funded by Google Summer of Code
  • Adding static multi-versioning capabilities to GCC and per-function fine-grain optimization for run-time adaptation
  • Associated publications [P30, P32]
[ Google Summer of Code description page ]
[Q10] 2009-04 - 2009-08 Postdoctoral researcher Cosmin Oancea in INRIA Saclay (France)
  • Exploratory research on program memory behavior characterization and run-time adaptation
[Q11] 2009-01 - 2009-06 Expert engineer Zbigniew Chamski in INRIA Saclay (France) funded by EU MILEPOST project
  • Moving Interactive Compilation Interface (event-based plugin framework) to mainline GCC
  • Publication [P26]
[Q12] 2008-03 - 2008-08 Postdoctoral researcher Abid Muslim Malik in INRIA Saclay (France) funded by EU MILEPOST project
  • Exploratory research on semantic program feature analysis to improve optimization predictions in MILEPOST GCC
[Q13] 2008-01 - 2008-02 MS student Lianjie Luo in Institute of Computing Technology (China)
  • Extending my previous work on statically enabling dynamic optimization [P57] with minimal representative sets of versions
  • Associated publication [P37]
[Q14] 2007-03 - 2007-04 PhD student Victor Jimenez in UPC (Spain) visiting INRIA Saclay (France) funded by EU HiPEAC fellowship
  • Extending my previous work on statically enabling dynamic optimization [P57] with predictive scheduling for heterogeneous architectures [P39]
  • Associated publication [P39]
[Q15] 2006-09 - 2008-08 PhD student Cupertino Miranda in INRIA Saclay (France)
  • Helping to develop Interactive Compilation Interface for MILEPOST GCC
  • Associated publications [P41, P42, P44, P45]
[Q16] 2006-09 - 2007-08 PhD student Piotr Lesnicki in INRIA Saclay (France)
  • Co-advising with Albert Cohen and Olivier Temam to extending my previous work on statically enabling dynamic optimization [P57] for split-compilation
  • Associated publication [P48]
[Q17] 2003-06 - 2003-08 MS student Edwin Bonilla in the University of Edinburgh (UK)
  • Exploring possibilities to use machine learning for fine-grain program optimization prediction (in GCC)
  • Publication [P53]

Organizing/chairing events

# Year CK Type Event
[E1] 2019-04 Artifact evaluation Gennady Pekhimenko (University of Toronto, Canada) and Grigori Fursin (dividiti, UK / cTuning foundation, France).
Artifact evaluation for SysML'19 , Stanford, CA, March 2019

[ AE website ][ SysML'19 website ]
[E2] 2018-10 Competition Quantum Computing Hack Day , London, UK, October 2018
  • Uses Collective Knowledge Framework to share resulting workflows and artifacts in a common format for a unified benchmarking and to let the community reuse and build upon best found solutions
[ Event website ][ CK workflow ][ CK quantum website ]
[E3] 2018-03 Competition Luis Ceze (University of Washington, USA), Natalie Enright Jerger (University of Toronto, Canada), Babak Falsafi (EPFL, Switzerland), Grigori Fursin (cTuning foundation, France / dividiti, UK), Anton Lokhmotov (dividiti, UK), Thierry Moreau (University of Washington, USA), Adrian Sampson (Cornell University, USA) and Phillip Stanley Marbell (University of Cambridge, UK).
ReQuEST: 1st open, reproducible and Pareto efficient SW/HW co-design competition for deep learning (speed, accuracy, costs) co-located with ACM ASPLOS 2018 , Williamsburg, VA, USA, March 2018
  • Sponsored by ACM, cTuning foundation and dividiti
  • Uses Collective Knowledge Framework to share resulting workflows and artifacts in a common format for a unified and unbiased benchmarking and to let the community reuse and build upon best found solutions
  • Supports our new publication model in computer engineering where all research artifacts (tools, benchmarks, datasets, models) are continuously shared and validated by the community [P14,E16,E17,E15,E12,E13]
  • Related publications [P9, P8, P14, P11, P15, P35]
[ Website ][ ACM proceedings ][ Online results report (proceedings front matter) ][ All reusable CK workflows and components ][ Reddit discussion ][ Collective Knowledge Framework used to put systems and ML researchers on the same page and share AI artifacts as customized, reproducible, reusable and optimized components with a unified JSON API ][ Live ReQuEST scoreboard with reproducible results ]
[E4] 2018-03 Panel Grigori Fursin (cTuning foundation, France / dividiti, UK), Thierry Moreau (University of Washington, USA), Hillery Hunter (IBM, USA), Yiran Chen (Duke University, USA), Charles Qi (Cadence, USA) and Tianqi Chen (University of Washington, USA).
Open panel and discussion on tackling complexity, reproducibility and tech transfer challenges in a rapidly evolving AI/ML/systems research co-located with ACM ReQuEST-ASPLOS'18 , Williamsburg, VA, USA, March 2018
  • Participants: Hillery Hunter (IBM, USA), Yiran Chen (Duke University, USA), Charles Qi (Cadence, USA), Tianqi Chen (University of Washington, USA)
[ Website ][ ACM proceedings (DOI) ]
[E5] 2018-03 Organizing committee Grigori Fursin (dividiti, UK / cTuning foundation, France).
Organizing Committee for ASPLOS'18 , Williamsburg, Virginia, USA, March 2018
[E6] 2018-02 Artifact evaluation Malavika Samak (MIT, USA) and Grigori Fursin (cTuning foundation, France / dividiti, UK).
Artifact evaluation for PPoPP'18 , Vienna, Austria, February 2018

[ AE website ][ PPoPP'18 website ]
[E7] 2018-02 Artifact evaluation Michel Steuwer (University of Glasgow, UK) and Grigori Fursin (cTuning foundation, France / dividiti, UK).
Artifact evaluation for CGO'18 , Vienna, Austria, February 2018

[ AE website ][ CGO'18 website ]
[E8] 2017-09 Artifact evaluation Grigori Fursin (dividiti, UK / cTuning foundation, France) and Bruce Childers (University of Pittsburgh, USA).
Artifact evaluation for PACT'17 , Portland, Oregon, September 2017

[ AE website ][ PACT'17 website ]
[E9] 2017-02 Artifact evaluation Wonsun Ahn (University of Pittsburgh, USA), Bruce Childers (University of Pittsburgh, USA) and Grigori Fursin (cTuning foundation, France / dividiti, UK).
Artifact evaluation for PPoPP'17 , Austin, Texas, USA, February 2017

[ AE website ][ PPoPP'17 website ]
[E10] 2017-02 Artifact evaluation Joseph Devietti (University of Pennsylvania, USA), Bruce Childers (University of Pittsburgh, USA) and Grigori Fursin (dividiti, UK / cTuning foundation, France).
Artifact evaluation for CGO'17 , Austin, Texas, USA, February 2017

[ AE website ][ CGO'17 website ]
[E11] 2016-09 Artifact evaluation Zheng Wang (Lancaster University, UK), Hugh Leather (University of Edinburgh, UK), Bruce Childers (University of Pittsburgh, USA) and Grigori Fursin (dividiti, UK / cTuning foundation, France).
Artifact evaluation for PACT'16 , Haifa, Israel, September 2016

[ AE website ][ PACT'16 website ]
[E12] 2016-03 Artifact evaluation Grigori Fursin (dividiti, UK / cTuning foundation, France) and Bruce Childers (University of Pittsburgh, USA).
Artifact evaluation for CGO'16 , Barcelona, Spain, March 2016

[ AE website ][ CGO'16 website ]
[E13] 2016-03 Artifact evaluation Grigori Fursin (dividiti, UK / cTuning foundation, France) and Bruce Childers (University of Pittsburgh, USA).
Artifact evaluation for PPoPP'16 , Barcelona, Spain, March 2016

[ AE website ][ PPoPP'16 website ]
[E14] 2016-01 Workshop Grigori Fursin (cTuning foundation, France / dividiti, UK) and Christophe Dubach (University of Edinburgh, UK).
6th International workshop on Adaptive Self-Tuning Computing Systems (ADAPT) with a new open publication model and public Reddit discussions co-located with HiPEAC 2016 , Prague, Czech Republic, January 2016
  • Keynote: "Benchmarks vs The Zombie Apocalypse: A comparison" by Ed Plowman (Director Performance Analysis Strategy, ARM, UK)
  • Authors of several articles have shared their artifacts in our open Collective Knowledge format
  • Sponsored by cTuning foundation, France and dividiti, UK
[ Website ][ Public discussion of submissions ]
[E15] 2015-11 Artifact evaluation Bruce Childers (University of Pittsburgh, USA), Grigori Fursin (dividiti, UK / cTuning foundation, France), Shriram Krishnamurthi (Brown University, USA) and Andreas Zeller (Saarland University, Germany).
Dagstuhl Perspectives Workshop: Artifact Evaluation for Publications , Dagstuhl, Germany, November 2015
[E16] 2015-02 Artifact evaluation Grigori Fursin (INRIA / cTuning foundation, France) and Bruce Childers (University of Pittsburgh, USA).
Artifact evaluation for PPoPP'15 , San Francisco Bay Area, CA, USA, Feburary 2015

[ AE website ][ PPoPP'15 website ]
[E17] 2015-02 Artifact evaluation Grigori Fursin (INRIA / cTuning foundation, France) and Bruce Childers (University of Pittsburgh, USA).
Artifact evaluation for CGO'15 , San Francisco Bay Area, CA, USA, Feburary 2015

[ AE website ][ CGO'15 website ]
[E18] 2015-01 Workshop Christophe Dubach (University of Edinburgh, UK) and Grigori Fursin (cTuning foundation, France).
5th International workshop on Adaptive Self-Tuning Computing Systems (ADAPT) with a special focus on reproducibility co-located with HiPEAC 2015 , Amsterdam, the Netherlands, January 2015
  • Keynote: "Accelerating Datacenter Services with Reconfigurable Logic" by Aaron Smith (Microsoft, USA)
  • Sponsored by Nvidia, USA and cTuning foundation, France
[ Website ][ Final program ][ One of ADAPT'15 papers has an interesting discussion on Slashdot ]
[E19] 2015-01 Special journal issue Alex Jones (University of Pittsburgh, USA) and Grigori Fursin (INRIA, France).
Special Issue on Reproducible Research Methodologies, IEEE Transactions on Emerging Topics in Computing (TETC)
  • Submission Deadline: Sept 1, 2014
  • Reviews Completed: Nov 1, 2014
  • Major Revisions Due (if Needed): Dec 15, 2014
  • Reviews of Revisions Completed (if Needed): Jan 3, 2015
  • Minor Revisions Due (if Needed): Jan 20, 2015
  • Notification of Final Acceptance: Jan 31, 2015
  • Publication Materials for Final Manuscripts Due: Feb 15, 2015
  • Publication date: 2nd Issue of 2015
[ IEEE TETC Website ][ Submission website ][ Flyer and CFP ][ Related OCCAM project ][ Related Collective Mind project ][ Related Collective Mind repository ]
[E20] 2014-06 Workshop Grigori Fursin (INRIA, France), Alex Jones (University of Pittsburgh, USA), Daniel Mosse (University of Pittsburgh, USA) and Bruce Childers (University of Pittsburgh, USA).
1st ACM SIGPLAN International Workshop on Reproducible Research Methodologies and New Publication Models (TRUST) co-located with PLDI 2014 , Edinburgh, UK, June 2014
  • Workshop Date: June 12, 2014
  • Abstract submission deadline: March 7, 2014 (Anywhere on Earth)
  • Paper submission deadline: March 14, 2014 (Anywhere on Earth)
  • Notification date: April 14, 2014
  • Final version deadline: May 2, 2014
  • Related ADAPT'14 panel on reproducible research methodologies and new publication models in computer engineering [E24]
[ Website ][ ACM DL ]
[E21] 2014-05 Panel Marisa Gil (BSC/UPC, Spain), Chris Fensch (University of Edinburgh, UK) and Grigori Fursin (INRIA, France).
Panel "Is Current Research on Heterogeneous HPC Platforms inline with Real-world Application needs?" co-located with HiPEAC Spring Computing Systems Week 2014 , BSC, Barcelona, Spain, May 2014
  • Participants: Paolo Faraboschi (HP Labs, Spain), Ana Lucia Varbanescu (University of Amsterdam, Netherlands), Mats Brorsson (KTH, Sweden), Pooyan Dadvand (CIMNE/UPC, Spain) and Paul Keir (CodePlay, UK);
[ Website and slides ][ CSW Website ]
[E22] 2014-02 Workshop Alex Jones (University of Pittsburgh, USA), Grigori Fursin (INRIA, France), Daniel Mosse (University of Pittsburgh, USA) and Bruce Childers (University of Pittsburgh, USA).
Workshop on Reproducible Research Methodologies (REPRODUCE) co-located with HPCA 2014 , Orlando, Florida, USA, February 2014
  • Submission Deadline: January 8, 2014
  • Notification of Acceptance: January 23, 2014
  • Final (Camera-ready) Manuscripts Due: January 30, 2014
  • Workshop Date: February 15, 2014
[ Website ]
[E23] 2014-01 Workshop Christophe Dubach (University of Edinburgh, UK) and Grigori Fursin (INRIA, France).
4th International workshop on Adaptive Self-Tuning Computing Systems (ADAPT) with a special focus on reproducibility co-located with HiPEAC 2014 , Vienna, Austria, January 2014
  • Keynote: "Towards Resource Management in Parallel Architectures Under the Hood" by Prof. Per Stenstrom (Chalmers University of Technology, Sweden)
  • Associated panel on reproducible research methodologies and new publication models in computer engineering [E24]
  • Shared research material and experimental results for 2 papers have been validated by our volunteers: Alberto Magni from the University of Edinburgh, UK and Sascha Hunold from Vienna University of Technology, Austria
  • Sponsored by Nvidia, USA
[ Website ][ Final program ][ Shared artifacts (paper 1) ][ Shared artifacts (paper 2) ]
[E24] 2014-01 Panel Grigori Fursin (INRIA, France), Alex Jones (University of Pittsburgh, USA), Daniel Mosse (University of Pittsburgh, USA) and Bruce Childers (University of Pittsburgh, USA).
Panel on reproducible research methodologies and new publication models in computer engineering co-located with ADAPT 2014 , Vienna, Austria, January 2014
  • Participants: Jack Davidson (University of Virginia / Co-Chair of ACM's Publication Board, USA); Lieven Eeckhout (Ghent University / Intel ExaScience Lab, Belgium); Sascha Hunold, Jesper Larsson Traff (Vienna University of Technology, Austria); Anton Lokhmotov (ARM, UK); Alex K.Jones, Daniel Mosse, Bruce Childers (University of Pittsburgh, USA); Grigori Fursin (INRIA, France);
  • Associated workshop [E23]
[ Website and slides ][ Brief blog note ]
[E25] 2013-05 Thematic session Grigori Fursin (INRIA, France).
Thematic session - making computer engineering a science co-located with ACM ECRC 2013 / HiPEAC computing week 2013 , Paris, France, May 2013
  • Keynote on "OCCAM: Open Curation for Computer Architecture Modeling" by Prof. Bruce Childers and Prof. Alex Jones (University of Pittsburgh, USA)
  • Talks by Vittorio Zaccaria (Politecnico di Milano, Italy), Christophe Guillon and Christian Bertin (STMicroelectronics, France), Christoph Reichenbach (Johann-Wolfgang Goethe Universitat Frankfurt, Germany)
  • Newsletter in [P18]
  • Sponsored by HiPEAC, EU
[ Website ][ Backup ]
[E26] 2013-01 Workshop Christophe Dubach (University of Edinburgh, UK) and Grigori Fursin (INRIA, France).
3rd International workshop on Adaptive Self-Tuning Computing Systems (ADAPT) co-located with HiPEAC 2013 , Berlin, Germany, January 2013
  • Keynote on "Autotuning Recursive Functions" by Prof. Markus Pueschel (ETHZ, Switzerland) sponsored by Microsoft Research
  • Introduction in [P19]
  • Sponsored by Microsoft Research, UK and Nvidia, USA
[ Website ][ ACM DL ]
[E27] 2012-04 Thematic session Grigori Fursin (INRIA, France).
Thematic session - "Collective characterization, optimization and design of computer systems" co-located with HiPEAC spring computing week 2012 , Goteborg, Sweden, April 2012
  • Talks by Marisa Gil (UPC, Spain), Lasse Natvig (NTNU, Norway), David Whalley (Florid State University, USA), Cristina Silvano (Politecnico di Milano, Italy)
  • Sponsored by HiPEAC, EU
[ Intro presentation (HAL) ][ Website ][ Backup ]
[E28] 2012-03 Workshop Grigori Fursin (Intel/CEA Exascale Lab, France), Jason Mars (University of Virginia, USA), Yuriy Kashnikov (Intel/CEA Exascale Lab, France) and Robert Hundt (Google, USA).
2nd International Workshop on Adaptive Self-Tuning Computing Systems for the Exaflop Era (EXADAPT) co-located with ASPLOS 2012 , London, UK, March 2012
  • Keynote on "Self-Tuning Bio-Inspired Massively-Parallel Computing" by Prof. Steve Furber (Manchester University, UK)
  • Introduction in [P24]
  • Associated panel [E29]
  • Sponsored by Google, USA
[ Website ][ ACM DL ]
[E29] 2012-03 Panel Grigori Fursin (INRIA, France).
Panel - Joint EXADAPT 2012 / GPGPU 2012 round table on "Leveraging GPUs and Self-Tuning Systems on the Road to Exascale" co-located with EXADAPT 2012 , London, UK, March 2012
  • Participants: Steve Furber (University of Manchester, UK), Anton Lokhmotov (ARM, UK), Paul Kelly (Imperial College London, UK)
  • Associated workshop [E28]
[ Website ]
[E30] 2011-11 BOF Grigori Fursin (INRIA, France) and Marie-Christine Sawley (Intel/CEA Exascale Lab, France).
BOF - Collaboratively mining rich information to prepare the Exascale challenges @ SC 2011 co-located with SuperComputing 2011 , Seattle, WA, USA, November 2011
  • Presenting and discussing long-term R&D of my team at Intel/CEA Exascale Lab, France
  • Continuation in [P17, P16]
[ Website ][ Backup ]
[E31] 2011-05 Workshop Grigori Fursin (Intel/CEA Exascale Lab, France), Jason Mars (University of Virginia, USA), Yuriy Kashnikov (Intel/CEA Exascale Lab, France) and Robert Hundt (Google, USA).
1st International ACM Workshop on Adaptive Self-Tuning Computing Systems for the Exaflop Era (EXADAPT) co-located with PLDI 2011 / FCRC 2011 , San Jose, USA, June 2011
  • Keynote on "Autotuning in the Exascale Era" by Prof. Katherine Yelick (LBNL and UC Berkeley, USA)
  • Introduction in [P27]
  • Sponsored by Google, USA and ACM, USA
[ Website ][ ACM DL ]
[E32] 2011-04 Workshop Grigori Fursin (Intel/CEA Exascale Lab, France) and John Cavazos (University of Delaware, USA).
5th International Workshop Statistical and Machine learning approaches to ARchitecture and compilaTion (SMART) co-located with CGO 2011 , Chamonix, France, April 2011
  • Keynote on "Automatic Performance Tuning and Machine Learning" by Prof. Markus Pueschel (ETHZ, Switzerland)
  • Chair: Prof. Francois Bodin (CAPS Entreprise, France)
  • Sponsored by CAPS Entreprise, France
[ Website ][ Final program ]
[E33] 2010-01 Tutorial Sid Touati (UVSQ, France) and Grigori Fursin (INRIA, France).
Tutorial - "Speedup-Test: Statistical Methodology to Evaluate Program Speedups and their Optimisation Techniques" co-located with HiPEAC 2010 , Pisa, Italy, January 2010

[ Website ]
[E34] 2010-01 Workshop Grigori Fursin (INRIA, France) and John Cavazos (University of Delaware, USA).
4th International Workshop Statistical and Machine learning approaches to ARchitecture and compilaTion (SMART) co-located with HiPEAC 2010 , Pisa, Italy, January 2010
  • Keynote on "Moving adaptation into individual optimizations" by Prof. Keith Cooper (Rice University, USA)
  • Chair: Prof. David Whalley (Florida State University, USA)
  • Sponsored by Intel/CEA Exascale Lab, France
[ Website ][ Final program ]
[E35] 2010-01 Panel Dorit Nuzman (IBM Haifa, Israel) and Grigori Fursin (INRIA, France).
2nd International Workshop on GCC Research Opportunities (GROW) co-located with HiPEAC 2010 , Pisa, Italy, January 2010
  • Keynote on "Using GCC as a toolbox for research: GCC plugins and whole-program compilation" by Diego Novillo (Google, Canada)
[ Website ]
[E36] 2009-06 Tutorial Grigori Fursin (INRIA, France).
cTuning tools tutorial on collaborative and reproducible program and architecture characterization and autotuning co-located with HiPEAC computing systems week , Infineon, Munich, Germany, June 2009
[E37] 2009-01 Workshop Grigori Fursin (INRIA, France) and John Cavazos (University of Delaware, USA).
3rd International Workshop Statistical and Machine learning approaches to ARchitecture and compilaTion (SMART) co-located with HiPEAC 2009 , Paphos, Cyprus, January 2009
  • Associated panel [E38]
  • Chair: Prof. David Padua (UIUC, USA)
[ Website ][ Final program ]
[E38] 2009-01 Panel Grigori Fursin (INRIA, France) and John Cavazos (University of Delaware, USA).
Panel - Can machine learning help to solve the multicore code generation issues? co-located with HiPEAC 2009 , Paphos, Cyprus, January 2009
  • Chair: Francois Bodin (CAPS-Enterprise, France); participants: Marcelo Cintra (University of Edinburgh, UK), Bilha Mendelson (IBM Haifa, Israel), Lawrence Rauchwerger (Texas A%M University, USA), Per Stenstrom (Chalmers University of Technology, Sweden)
  • Associated workshop [E37]
[ Website ][ Video ]
[E39] 2008-01 Workshop Grigori Fursin (INRIA, France) and John Cavazos (University of Delaware, USA).
2nd International Workshop Statistical and Machine learning approaches to ARchitecture and compilaTion (SMART) co-located with HiPEAC 2008 , Goteborg, Sweden, January 2008
  • Chair: Prof. Michael O'Boyle (University of Edinburgh, UK)
[ Website ]
[E40] 2007-01 Tutorial Albert Cohen (INRIA, France), Ayal Zaks (IBM Haifa, Israel), Dorit Nuzman (IBM Haifa, Israel), Diego Novillo (Red Hat, USA), Roberto Costa (STMicroelectronics, Italy), Grigori Fursin (INRIA, France) and Sebastian Pop (Ecole des Mines de Paris, France).
2nd HiPEAC GCC Tutorial: How To and Return on Experience co-located with HiPEAC 2007 , Ghent, Belgium, January 2007

[ Website ]
[E41] 2007-01 Workshop Grigori Fursin (INRIA, France) and John Cavazos (University of Delaware, USA).
1st International Workshop Statistical and Machine learning approaches to ARchitecture and compilaTion (SMART) co-located with HiPEAC 2007 , Ghent, Belgium, January 2007

[ Website ][ Final program ]
[E42] 2001-05 Workshop Michael O'Boyle (University of Edinburgh, UK) and Grigori Fursin (University of Edinburgh, UK).
9th Workshop on Compilers for Parallel Computers (CPC) , Edinburgh, UK, June 2001
  • Local organizers: Michael O'Boyle (University of Edinburgh, UK), Grigori Fursin (University of Edinburgh, UK)
[ Website ]

Publications
# Year CK Type Citation
[P1] 2021 Journal Grigori Fursin.
Collective Knowledge: organizing research projects as a database of reusable components and portable workflows with common APIs.
Accepted for Philosophical Transactions A of the Royal Society, March 2021

[CK bib / cM bib] [View doc from CK / cM ][ DOI ][ preprint ]
[P2] 2021 Technical report Rafael Ferreira da Silva, Henri Casanova, Kyle Chard, Dan Laney, Dong Ahn, Shantenu Jha, Carole A. Goble, Lavanya Ramakrishnan, Luc Peterson, Bjoern Enders, Douglas Thain, Ilkay Altintas, Yadu N. Babuji, Rosa M. Badia, Vivien Bonazzi, Taina Coleman, Michael R. Crusoe, Ewa Deelman, Frank Di Natale, Paolo Di Tommaso, Thomas Fahringer, Rosa Filgueira, Grigori Fursin, Alex Ganose, Bjorn Gruning, Daniel S. Katz, Olga Kuchar, Ana Kupresanin, Bertram Ludascher, Ketan Maheshwari, Marta Mattoso, Kshitij Mehta, Todd Munson, Jonathan Ozik, Tom Peterka, Loic Pottier, Tim Randles, Stian Soiland-Reyes, Benjamin Tovar, Matteo Turilli, Thomas D. Uram, Karan Vahi, Michael Wilde, Matthew Wolf and Justin M. Wozniak.
Workflows Community Summit: Bringing the Scientific Workflows Community Together.
arXiv:2103.09181, March 2021

[CK bib / cM bib] [View doc] [View doc from CK backup / cM backup ]
[P3] 2021 Journal Paolo Sylos Labini, Marco Cianfriglia, Damiano Perri, Grigori Fursin, Anton Lokhmotov, Cedric Nugteren, Bruno Carpentieri, Fabiana Zollo and Flavio Vella.
On the Anatomy of Predictive Models for Accelerating GPU Convolution Kernels and Beyond.
ACM Transactions on Architecture and Code Optimization (TACO), January 2021

[CK bib / cM bib] [View doc from CK / cM ][ ACM DL ]
[P4] 2020 Technical report Grigori Fursin.
The Collective Knowledge project: making ML models more portable and reproducible with open APIs, reusable best practices and MLOps.
arXiv:2006.07161, June 2020
  • I summarized my experience working with companies, universities and non-profits to bridge the growing gap between AI, ML and systems research and practice with the help of collaborative knowledge management, portable workflows, reusable artifacts and best practices, unified meta descriptions, open APIs, and reproducible experiments.
[CK bib / cM bib] [View doc] [View doc from CK backup / cM backup ][ Collective Knowledge repository with portable workflows, reusable artifacts and scoreboards for collaborative and reproducible experiments ][ Collective Knowledge portal ][ Collective Knowledge Framework ]
[P5] 2018 Technical report Luis Ceze, Natalie Enright Jerger, Babak Falsafi, Grigori Fursin, Anton Lokhmotov, Thierry Moreau, Adrian Sampson and Adrian Sampson.
Proceedings Front Matter: Introducing the 1st ACM ReQuEST Workshop/Tournament on Reproducible Software/Hardware Co-design of Pareto-Efficient Deep Learning.
ACM Digital Library, June 2018

[CK bib / cM bib] [View doc][ Proceedings front-matter ][ ReQuEST-ASPLOS'18 ACM proceedings ][ ReQuEST live scoreboard with SW/HW/AI/ML co-design configurations ]
[P6] 2018 Technical report Marco Cianfriglia, Flavio Vella, Cedric Nugteren, Anton Lokhmotov and Grigori Fursin.
A model-driven approach for a new generation of adaptive libraries.
arXiv tech.report 1806.07060, June 2018

[CK bib / cM bib] [View doc][ arXiv ]
[P7] 2018 Technical report Anton Lokhmotov, Nikolay Chunosov, Flavio Vella and Grigori Fursin.
Multi-objective autotuning of MobileNets across the full software/hardware stack.
Proceedings of the 1st on Reproducible Quality-Efficient Systems Tournament on Co-designing Pareto-efficient Deep Learning (ReQuEST-ASPLOS'18), June 2018

[CK bib / cM bib] [View doc][ ACM DL (DOI) ]
[P8] 2018 Technical report Grigori Fursin, Anton Lokhmotov, Dmitry Savenko and Eben Upton.
A Collective Knowledge workflow for collaborative research into multi-objective autotuning and machine learning techniques.
arXiv tech.report 1801.08024, January 2018

[CK bib / cM bib] [View doc][ arXiv ][ Interactive CK report ][ GitHub repository ][ FigShare CK archives ][ Live CK repository with latest crowd-tuning results ]
[P9] 2018 Technical report Thierry Moreau, Anton Lokhmotov and Grigori Fursin.
Introducing ReQuEST: an Open Platform for Reproducible and Quality-Efficient Systems-ML Tournaments.
arXiv tech.report 1801:06378, January 2018

[CK bib / cM bib] [View doc][ arXiv ][ ReQuEST tournament web page ][ CK live scoreboard with SW/HW/AI/ML co-design configurations ]
[P10] 2016 Workshop Anton Lokhmotov and Grigori Fursin.
Nearly Everything You Need to Know About Optimizing Convolutional Neural Networks on Embedded Platforms with OpenCL.
International Workshop on OpenCL (IWOCL), Vienna, Austria, April 2016
  • Related Collective Knowledge infrastructure and repository (CK) [S1]
[CK bib / cM bib] [View doc][ Slides ][ Collective Knowledge Live demo repository ][ Collective Knowledge Framework ][ Android app for CK for collaborative program optimization and machine learning using mobile devices ][ Community-driven gemmbench (OpenCL) ]
[P11] 2016 Conference Grigori Fursin, Anton Lokhmotov and Ed Plowman.
Collective Knowledge: towards R&D sustainability.
DATE 2016: Design, automation and test in Europe, Dresden, Germany, March 2016
  • Partially funded by [F3]
  • This work summarizes our long-term vision on enabling systematic benchmarking and cost-aware computer engineering autotuning using methodology from physics, biology and other natural sciences
  • Related Collective Knowledge infrastructure and repository (CK) [S2]
[CK bib / cM bib][ Collective Knowledge Framework ][ Collective Knowledge Live demo repository ][ Community-driven gemmbench (OpenCL) ][ Related artifact evaluation for CGO and PPoPP'16 ][ Related lecture at the University of Manchester ]
[P12] 2016 Workshop Bruce Childers, Grigori Fursin, Shriram Krishnamurthi and Andreas Zeller.
Artifact Evaluation for Publications.
Dagstuhl Perspectives Workshop 15452, Prague, Czech Republic, March 2016
  • Related Dagstuhl Perspectives Workshop 15452 [E15]
[CK bib / cM bib][ Official doc and BibTex reference ]
[P13] 2016 Workshop Anton Lokhmotov and Grigori Fursin.
Collaborative design and optimization using Collective Knowledge.
MULTIPROG 2016: Programmability and Architectures for Heterogeneous Multicores, Prague, Czech Republic, January 2016
  • Related Collective Knowledge infrastructure and repository (CK) [S1]
[CK bib / cM bib] [View doc][ Collective Knowledge Live demo repository ][ Collective Knowledge Framework ][ Android app for CK for collaborative program optimization and machine learning using mobile devices ][ Community-driven gemmbench (OpenCL) ][ Related artifact evaluation for CGO and PPoPP'16 ][ Related lecture at the University of Manchester ]
[P14] 2014 Workshop Grigori Fursin and Christophe Dubach.
Experience report: community-driven reviewing and validation of publications.
Proceedings of the 1st Workshop on Reproducible Research Methodologies and New Publication Models in Computer Engineering (TRUST 2014) co-located with PLDI 2014, Edinburgh, UK, June 2014
  • We propose a new and open publication model for reproducible research where articles, experiments and artifacts are reviewed by the community based on our practical related experience during MILEPOST, cTuning and Collective Mind projects since 2008
[CK bib / cM bib] [View doc] [View doc] [View doc] [View doc] [View doc from CK backup / cM backup ]
[P15] 2014 Extended abstract Grigori Fursin.
Crowdsourcing autotuning: challenges and possible solutions.
Extended abstract at Dagstuhl Seminar 13401, Dagstuhl, Germany, January 2014

[CK bib / cM bib] [View doc] [View doc from CK backup / cM backup ]
[P16] 2013 Workshop Abdul Wahid Memon and Grigori Fursin.
Crowdtuning: systematizing autotuning using predictive modeling and crowdsourcing.
Proceedings of the Application Autotuning for HPC mini-symposium co-located with PARCO 2013, Munich, Germany, September 2013

[CK bib / cM bib] [View doc] [View doc]
[P17] 2013 Technical report Grigori Fursin.
Collective Mind: cleaning up the research and experimentation mess in computer engineering using crowdsourcing, big data and machine learning.
INRIA technical report HAL-00850880, France, 2013
  • Extended journal version: [P15]
  • This work summarizes my long-term vision on collaborative, systematic and reproducible benchmarking, optimization and co-design of computer systems across all software and hardware layers using public Collective Mind repository of knowledge, common plugin-based autotuning framework, big data, predictive analytics (machine learning, data mining, statistical analysis, feature detection), crowdsourcing and collective intelligence
  • This work extends my previous article [P35]
  • Should be publicly available at some point in autumn, 2014
  • Related Collective Mind infrastructure and repository [S3]
  • This work supports my initiative on open research and publication model where all experimental results and related material is continuously shared, validated and improved by the community [P14]. To set up an example, I continue sharing all benchmarks, datasets, tools, models and experimental results in Collective Mind repository (c-mind.org/repo)
[CK bib / cM bib] [View doc] [View doc] [View doc from CK backup / cM backup ]
[P18] 2013 Newsletter Grigori Fursin.
HiPEAC Thematic Session on "Making Computer Engineering a science": cleaning up the mess.
HiPEAC newsletter 35, 2013
  • Introducing open-source Collective Mind Framework (plugin-based knowledge management system) [S3] and public repository [R3] to start collaborative systematization of computer engineering and initiate new publication model where all research artifacts are shared, validated and extended by the community [M2]
[CK bib / cM bib] [View doc from CK / cM ]
[P19] 2013 Introduction Christophe Dubach and Grigori Fursin.
Introducing 3rd International Workshop on Adaptive Self-Tuning Computing Systems.
ACM Digital Library, 2013

[CK bib / cM bib] [View doc] [View doc from CK backup / cM backup ]
[P20] 2013 Poster Grigori Fursin.
Poster: cTuning.org: novel collaborative methodology, publication model, framework and repository to crowdsource autotuning.
HiPEAC conference poster session, Berlin, Germany, January 2013

[CK bib / cM bib] [View doc from CK / cM ]
[P21] 2012 Extended abstract Grigori Fursin.
cTuning.org: Novel Extensible Methodology, Framework and Public Repository to Collaboratively Address Exascale Challenges.
Extended abstract at SuperComputing Companion (SC), pages 1401-1402, Salt Lake City, Utah,USA, November 2012

[CK bib / cM bib] [View doc] [View doc from CK backup / cM backup ]
[P22] 2012 Poster Grigori Fursin.
cTuning.org: novel extensible methodology, framework and public repository to collaboratively address Exascale challenges.
Poster at SuperComputing Companion (SC), pages 1403, Salt Lake City, Utah,USA, 2012

[CK bib / cM bib] [View doc] [View doc from CK backup / cM backup ]
[P23] 2012 Journal Yang Chen, Shuangde Fang, Yuanjie Huang, Lieven Eeckhout, Grigori Fursin, Olivier Temam and Chengyong Wu.
Deconstructing iterative optimization.
ACM Transactions on Architecture and Code Optimization (TACO), Volume 9, Number 3, pages 21:1-21:30, September 2012

[CK bib / cM bib] [View doc] [View doc from CK backup / cM backup ]
[P24] 2012 Introduction Grigori Fursin, Yuriy Kashnikov, Jason Mars and Robert Hundt.
Introducing 2nd International Workshop on Adaptive Self-Tuning Computing Systems for the Exaflop Era (EXADAPT).
ACM Digital Library, 2012

[CK bib / cM bib] [View doc] [View doc from CK backup / cM backup ]
[P25] 2012 Poster Grigori Fursin.
Poster: cTuning.org: Collaborative initiative to create open-source repository and tools to share and reuse knowledge about designs and optimizations of computer systems.
HiPEAC conference poster session, Paris, France, January 2012

[CK bib / cM bib] [View doc from CK / cM ]
[P26] 2011 Journal 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, Christopher Williams and Michael O'Boyle.
Milepost GCC: Machine Learning Enabled Self-tuning Compiler.
International Journal of Parallel Programming, pages 296-327, June 2011
  • Implements concept [M4]
  • Funded by [F9, F6]
  • Collaboration with IBM (Israel), University of Edinburgh (UK), ARC (now Synopsys, UK), CAPS Entreprise (France), and ICT (China)
  • MILEPOST GCC software [S10]; most of technology is now available in mainline GCC and is being added to major commercial compilers
  • cTuning CC software [S9]
  • cTuning repository [R5]
  • Award [A4]
  • Considered by IBM to be the first practical machine-learning based compiler in the world (IBM press-release [P34])
  • All benchmarks, datasets, tools, models and experimental results have been released to public for collaborative validation and extension!
[CK bib / cM bib] [View doc] [View doc from CK backup / cM backup ]
[P27] 2011 Introduction Grigori Fursin, Yuriy Kashnikov, Jason Mars and Robert Hundt.
Introducing ACM SIGPLAN International Workshop on Adaptive Self-Tuning Computing Systems for the Exaflop Era (EXADAPT).
ACM Digital Library, 2011

[CK bib / cM bib] [View doc] [View doc from CK backup / cM backup ]
[P28] 2010 Journal Grigori Fursin and Olivier Temam.
Collective optimization: A practical collaborative approach.
ACM Transactions on Architecture and Code Optimization (TACO), Volume 7, Number 4, pages 20:1-20:29, December 2010
  • Implements concept [M5]
  • Award [A4]
  • Extended paper [P38]
  • All benchmarks, datasets, tools, models and experimental results have been released in cTuning public optimization repository of knowledge [R5] for collaborative validation and extension!
  • Continues in [M2]
[CK bib / cM bib] [View doc] [View doc from CK backup / cM backup ]
[P29] 2010 Poster Souad Koliai, Grigori Fursin, Tipp Moseley and William Jalby.
DECAN: Decremental Performance Analysis Tool via Binary Patching.
Poster at the Workshop on Languages and Compilers for Parallel Computing (LCPC), USA, 2010
  • Extension of publication [P59]
  • Extension of concept [M8]
  • Discontinused for Alchemist plugin within Collective Mind Framework [S3, S5]
[CK bib / cM bib] [View doc from CK / cM ]
[P30] 2010 Conference Yang Chen, Yuanjie Huang, Lieven Eeckhout, Grigori Fursin, Liang Peng, Olivier Temam and Chengyong Wu.
Evaluating iterative optimization across 1000 datasets.
Proceedings of the 2010 ACM SIGPLAN conference on Programming language design and implementation (PLDI), pages 448-459, Toronto, Canada, June 2010 (acceptance rate: 20% (41/204))
  • Extension of [P49]
  • Being added to public repository of knowledge [R3]
[CK bib / cM bib] [View doc] [View doc from CK backup / cM backup ]
[P31] 2010 Conference Mircea Namolaru, Albert Cohen, Grigori Fursin, Ayal Zaks and Ari Freund.
Practical aggregation of semantical program properties for machine learning based optimization.
Proceedings of the International Conference on Compilers, Architectures and Synthesis for Embedded Systems (CASES'10), pages 197-206, Scottsdale, Arizona, USA, October 2010 (acceptance rate=29%)

[CK bib / cM bib] [View doc from CK / cM ]
[P32] 2010 Workshop Yuanjie Huang, Liang Peng, Chengyong Wu, Yuriy Kashnikov, Joern Renneke and Grigori Fursin.
Transforming GCC into a research-friendly environment: plugins for optimization tuning and reordering, function cloning and program instrumentation.
2nd International Workshop on GCC Research Opportunities (GROW), co-located with HiPEAC'10 conference, Pisa, Italy, January 2010 (acceptance rate: 57% (8/14))

[CK bib / cM bib] [View doc] [View doc from CK backup / cM backup ]
[P33] 2009 Conference Christophe Dubach, Timothy M. Jones, Edwin Bonilla, Grigori Fursin and Michael O'Boyle.
Portable compiler optimisation across embedded programs and microarchitectures using machine learning.
Proceedings of the 42nd Annual IEEE/ACM International Symposium on Microarchitecture (MICRO), pages 78-88, New York, NY, USA, December 2009 (acceptance rate: 25% (52/209))
  • HiPEAC paper award [A6]
  • Includes concept [M8]
  • Christophe Dubach received BCS/CPHC Distinguished Dissertation Award'09 for his related thesis "Using Machine-Learning to Efficiently Explore the Architecture/Compiler Co-Design Space" supervised by Prof. Michael O'Boyle.
[CK bib / cM bib] [View doc from CK / cM ]
[P34] 2009 Press release World's First Intelligent, Open Source Compiler Provides Automated Advice on Software Code Optimization (IBM Research and European Union Provide Software Developers with Performance Gains and Faster Time-To-Market).
IBM MILEPOST project press release, Haifa, Israel and Armonk, NY, USA, June 2009

[CK bib / cM bib] [View doc]
[P35] 2009 Workshop Grigori Fursin.
Collective Tuning Initiative: automating and accelerating development and optimization of computing systems.
Proceedings of the GCC Developers' Summit, Montreal, Canada, June 2009
  • Introduced concept of crowd-tuning using public repositories of knowledge, autotuning, machine learning and crowd-sourcing, and new publication model where results are continuously validated and extended by the community [M5]
  • Award [A4]
  • MILEPOST GCC software [S10]; most of technology is now available in mainline GCC and is being added to major commercial compilers
  • cTuning CC software [S9]
  • cTuning public optimization repository of knowledge [R5]
  • All benchmarks, datasets, tools, models and experimental results have been released to public for collaborative validation and extension! See our new research and publication proposal [P14]!
  • Continues in [M2, P17, P15]
[CK bib / cM bib] [View doc] [View doc from CK backup / cM backup ]
[P36] 2009 Conference John Thomson, Michael O'Boyle, Grigori Fursin and Björn Franke.
Reducing training time in a one-shot machine learning-based compiler.
Proceedings of the 22nd international conference on Languages and Compilers for Parallel Computing (LCPC), Newark, DE, USA, October 2009

[CK bib / cM bib] [View doc from CK / cM ]
[P37] 2009 Workshop Lianjie Luo, Yang Chen, Chengyong Wu, Shun Long and Grigori Fursin.
Finding representative sets of optimizations for adaptive multiversioning applications.
3rd Workshop on Statistical and Machine Learning Approaches Applied to Architectures and Compilation (SMART'09), co-located with HiPEAC'09 conference, Paphos, Cyprus, 2009 (acceptance rate=62% (8/13))
[CK bib / cM bib] [View doc] [View doc from CK backup / cM backup ]
[P38] 2009 Conference Grigori Fursin and Olivier Temam.
Collective Optimization.
Proceedings of the 4th International Conference on High Performance Embedded Architectures and Compilers (HiPEAC), pages 34-49, Paphos, Cyprus, January 2009 (acceptance rate: 28% (27/97))
  • Implements concept [M5]
  • Considerably extended in journal version [P28]
[CK bib / cM bib] [View doc from CK / cM ]
[P39] 2009 Conference Víctor J. Jiménez, Lluís Vilanova, Isaac Gelado, Marisa Gil, Grigori Fursin and Nacho Navarro.
Predictive Runtime Code Scheduling for Heterogeneous Architectures.
Proceedings of the 4th International Conference on High Performance Embedded Architectures and Compilers (HiPEAC), pages 19-33, Paphos, Cyprus, January 2009 (acceptance rate=28% (27/97))
  • I initiated collaborative HiPEAC project and obtained funding [F7] to extend [P57, M6]
  • Similar approaches for gluing/adapting applications for heterogeneous architectures are now used in Intel's Qilin and in CAPS Entreprise's HMPP frameworks
  • Being extended in [P17, M2]
[CK bib / cM bib] [View doc from CK / cM ]
[P40] 2009 Journal Shun Long and Grigori Fursin.
Systematic search within an optimisation space based on Unified Transformation Framework.
International Journal of Computational Science and Engineering (IJCSE), Volume 4, Number 2, pages 102-111, July 2009

[CK bib / cM bib] [View doc from CK / cM ]
[P41] 2008 Poster Grigori Fursin, Cupertino Miranda, Olivier Temam, Mircea Namolaru, Elad Yom-Tov, Ayal Zaks, Bilha Mendelson, Phil Barnard, Elton Ashton, Eric Courtois, Francois Bodin, Edwin Bonilla, John Thomson, Hugh Leather, Chris Williams and Michael O'Boyle.
MILEPOST GCC: machine learning based research compiler.
Poster at ACACES summer school, Italy, 2008

[CK bib / cM bib] [View doc from CK / cM ]
[P42] 2008 Workshop Grigori Fursin, Cupertino Miranda, Olivier Temam, Mircea Namolaru, Elad Yom-Tov, Ayal Zaks, Bilha Mendelson, Phil Barnard, Elton Ashton, Eric Courtois, Francois Bodin, Edwin Bonilla, John Thomson, Hugh Leather, Chris Williams and Michael O'Boyle.
MILEPOST GCC: machine learning based research compiler.
Proceedings of the GCC Developers' Summit, Ottawa, Canada, June 2008
  • Implements concept [M4]
  • Funded by EU MILEPOST project [J12, F9]
  • Considerably extended in journal version [P26]
[CK bib / cM bib] [View doc] [View doc from CK backup / cM backup ]
[P43] 2007 Workshop Veerle Desmet, Grigori Fursin, Sylvain Girbal and Olivier Temam.
Leveraging Modular Simulation for Systematic Design Space Exploration.
4th HiPEAC Industrial Workshop on Compilers and Architectures organized by ARM Ltd., Cambridge, UK, November 2007

[CK bib / cM bib]
[P44] 2007 Poster Grigori Fursin, Cupertino Miranda, Sebastian Pop and Albert Cohen.
Enabling Interactivity in GCC for Fine-Grain Optimizations.
Poster at ACACES summer school, Italy, 2007

[CK bib / cM bib] [View doc from CK / cM ]
[P45] 2007 Workshop Grigori Fursin, Cupertino Miranda, Sebastian Pop, Albert Cohen and Olivier Temam.
Practical Run-time Adaptation with Procedure Cloning to Enable Continuous Collective Compilation.
Proceedings of the GCC Developers' Summit, Ottawa, Canada, June 2008
  • Moving concept of statically enabling dynamic optimizations [M6] to mainline GCC
[CK bib / cM bib] [View doc from CK / cM ]
[P46] 2007 Conference Christophe Dubach, John Cavazos, Björn Franke, Grigori Fursin, Michael O'Boyle and Olivier Temam.
Fast compiler optimisation evaluation using code-feature based performance prediction.
Proceedings of the 4th international conference on Computing Frontiers, pages 131-142, Ischia, Italy, May 2007 (acceptance rate=50% (28/56))

[CK bib / cM bib] [View doc from CK / cM ]
[P47] 2007 Conference John Cavazos, Grigori Fursin, Felix Agakov, Edwin Bonilla, Michael O'Boyle and Olivier Temam.
Rapidly Selecting Good Compiler Optimizations using Performance Counters.
Proceedings of the International Symposium on Code Generation and Optimization (CGO), pages 185-197, San Jose, USA, March 2007
  • We added hardware counters to [S15, S11, R6] and used PCA to improve machine-learning based optimization prediction
  • Our paper received ACM CGO'17 test of time award: A1
[CK bib / cM bib] [View doc from CK / cM ]
[P48] 2007 Workshop Piotr Lesnicki, Albert Cohen, Grigori Fursin, Marco Cornero, Andrea Ornstein and Erven Rohou.
Split Compilation: an Application to Just-in-Time Vectorization.
International Workshop on GCC for Research in Embedded and Parallel Systems (GREPS'07) co-located with PACT'07, Brasov, Romania, September 2007

[CK bib / cM bib] [View doc] [View doc from CK backup / cM backup ]
[P49] 2007 Conference Grigori Fursin, John Cavazos, Michael O'Boyle and Olivier Temam.
MiDataSets: creating the conditions for a more realistic evaluation of Iterative optimization.
Proceedings of the 2nd international conference on High performance embedded architectures and compilers (HiPEAC), pages 245-260, Ghent, Belgium, January 2007 (acceptance rate=29%)
  • Preparing R&D on collective optimization [M5, M4, M2]
  • Released in public repositories of knowledge [R5, R3]
[CK bib / cM bib] [View doc from CK / cM ]
[P50] 2007 Workshop Grigori Fursin and Albert Cohen.
Building a Practical Iterative Interactive Compiler.
1st Workshop on Statistical and Machine Learning Approaches Applied to Architectures and Compilation (SMART'07), colocated with HiPEAC 2007 conference, Ghent, Belgium, January 2007 (acceptance rate=58% (7/12))
  • In this paper I introduced a novel concept to convert hardwired blackboxed production compilers into interactive research toolsets [M7]
  • Extended journal version in [P26]
  • Associated public software [S17, S10]
  • Discontinued for OpenME interface [S4, P17]
  • Now available in mainline GCC since version 4.6 [F6]
[CK bib / cM bib] [View doc from CK / cM ]
[P51] 2007 Conference Shun Long, Grigori Fursin and Björn Franke.
A cost-aware parallel workload allocation approach based on machine learning techniques.
Proceedings of the 2007 IFIP international conference on Network and Parallel Computing (NPC), pages 506-515, Dalian, China, September 2007

[CK bib / cM bib] [View doc] [View doc from CK backup / cM backup ]
[P52] 2006 Conference John Cavazos, Christophe Dubach, Felix Agakov, Edwin Bonilla, Michael O'Boyle, Grigori Fursin and Olivier Temam.
Automatic performance model construction for the fast software exploration of new hardware designs.
Proceedings of the International Conference on Compilers, Architecture and Synthesis for Embedded Systems (CASES), pages 24-34, Seoul, Korea, October 2006 (acceptance rate=41% (41/100))
  • Finalist best paper award
[CK bib / cM bib] [View doc from CK / cM ]
[P53] 2006 Conference Felix Agakov, Edwin Bonilla, John Cavazos, Björn Franke, Grigori Fursin, Michael O'Boyle, John Thomson, Mark Toussaint and Christopher K. I. Williams.
Using Machine Learning to Focus Iterative Optimization.
Proceedings of the International Symposium on Code Generation and Optimization (CGO), pages 295-305, New York, NY, USA, March 2006 (acceptance rate=36% (29/80))
  • Best presentation award
  • Concept [M4]
  • Extended journal version in [P26]
[CK bib / cM bib] [View doc from CK / cM ]
[P54] 2006 Journal Grigori Fursin, Albert Cohen, Michael O'Boyle and Olivier Temam.
Quick and Practical Run-Time Evaluation of Multiple Program Optimizations.
Transactions on High-Performance Embedded Architectures and Compilers, Volume 1, pages 13-31, 2006
  • Extended version of [P57]
[CK bib / cM bib] [View doc] [View doc from CK backup / cM backup ]
[P55] 2005 Conference Björn Franke, Michael O'Boyle, John Thomson and Grigori Fursin.
Probabilistic source-level optimisation of embedded programs.
Proceedings of the 2005 ACM SIGPLAN/SIGBED conference on Languages, Compilers, and Tools for Embedded Systems (LCTES), pages 78-86, Chicago, IL, USA, June 2005 (acceptance rate=26% (25/95))
  • Introducing probabilistic approach for adaptive exploration of large optimization spaces (effectively reducing tuning dimensions)
[CK bib / cM bib] [View doc from CK / cM ]
[P56] 2005 Workshop Shun Long and Grigori Fursin.
A Heuristic Search Algorithm Based on Unified Transformation Framework.
Proceedings of the 7th International Workshop on High Performance Scientific and Engineering Computing (HPSEC), pages 137-144, Oslo, Norway, June 2005
  • Our first experiments on program autotuning using polyhedral transformations (iterative optimization using polyhedral model)
[CK bib / cM bib] [View doc from CK / cM ]
[P57] 2005 Conference Grigori Fursin, Albert Cohen, Michael O'Boyle and Olivier Temam.
A practical method for quickly evaluating program optimizations.
Proceedings of the First International Conference on High Performance Embedded Architectures and Compilers (HiPEAC), pages 29-46, Barcelona, Spain, November 2005 (acceptance rate=20% (17/84))
  • Highest ranked paper [A7]
  • In this paper I introduced a novel concept to statically enable run-time optimizations and self-tuning binaries through code cloning and integrated low-overhead program/system behaviour monitoring plugin that can evaluate or select multiple optimization online using program phases [M6]. It has been referenced in patents and extended in academia and industry. We utilized Interactive Compilation Interface for PathScale compiler with loop vectorization, tiling, unrolling, interchange, fission/fusion, pipelining, prefetching and array padding to make static binaries adaptable and reactive to various environments and run-time behaviour to improve execution time and power consumption.
  • Presented ideas are now being used and extended by Google for data centers (cloud computing) to build adaptive applications and save energy [F6]
  • Funded by [A8, F6]
  • Associated public software [S14, S15, S17]
  • Multi-versioning support for adaptive appliations is now available in mainline GCC since version 4.8 [F6]
[CK bib / cM bib] [View doc] [View doc from CK backup / cM backup ]
[P58] 2004 PhD thesis Grigori Fursin.
Iterative Compilation and Performance Prediction for Numerical Applications.
PhD thesis, University of Edinburgh, UK, May 2004
  • Based on publications [P63, P61, P59]
  • PhD degree [Z6]
  • Concept [M9]
  • Presented ideas are now being used and extended in Intel Exascale Lab (France) [I2, P29]
  • Associated public software [S20]
[CK bib / cM bib] [View doc] [View doc from CK backup / cM backup ]
[P59] 2004 Journal Grigori Fursin, Michael O'Boyle, Olivier Temam and Gregory Watts.
A fast and accurate method for determining a lower bound on execution time.
Concurrency: Practice and Experience, Volume 16, Number 2-3, pages 271-292, January 2004
  • Concept [M8]
  • Extended version of publication [P63]
  • Associated public software [S19]
  • Extended in PhD thesis [P58] and DECAN framework [P29]
[CK bib / cM bib] [View doc] [View doc from CK backup / cM backup ]
[P60] 2003 Poster Grigori Fursin.
Poster: Iterative Compilation and Performance Prediction.
Division of Informatics, University of Edinburgh, UK, 2003

[CK bib / cM bib] [View doc from CK / cM ]
[P61] 2002 Workshop Grigori Fursin, Michael O'Boyle and Peter Knijnenburg.
Evaluating Iterative Compilation.
Proceedings of the 15th Workshop on Languages and Compilers for Parallel Computing (LCPC), pages 305-315, College Park, MD, USA, 2002
  • In this paper I introduced a concept of empirical optimization for large applications to automatically adapt them to a given hardware using several basic search strategies including random and hill-climbing. This approach considerably outperformed state-of-art compilers on Intel, Alpha and several other popular architectures for several large SPEC applications. This technique has also laid foundations for further research on systematic program and architecture optimization and co-design using statistical analysis, machine learning and run-time adaptation [M5, M4, M2]
  • Associated public software [S19]
[CK bib / cM bib] [View doc] [View doc from CK backup / cM backup ]
[P62] 2001 Technical report Grigori Fursin, Michael O'Boyle, Olivier Temam and Gregory Watts.
A Fast and Accurate Evaluation of Memory Performance Upper-Bound.
ESPRIT project No 24942 technical report, 2001
  • Part of EU FP5 MHAOTEU project [J17]
  • Associated public software [S19]
[CK bib / cM bib]
[P63] 2001 Workshop Grigori Fursin, Michael O'Boyle, Olivier Temam and Gregory Watts.
Fast and Accurate Evaluation of Memory Performance Upper-Bound.
Proceedings of the 9th Workshop on Compilers for Parallel Computers (CPC), pages 163-172, Edinburgh, UK, 2001
  • In this paper I introduced a novel, simple and fast approach to detect program performance anomalies or CPU/memory bounds via semantically non-equivalent assembler patching [M8]. We add or remove various assembler instructions to convert array accesses to scalars in various ways without preserving the semantics of the code while avoiding code crashing to be able to directly compare original and transformed programs. This technique does not need any slow simulation and proved to be realistic particularly on out-of-order processors where hardware counters can be totally misleading. This technique also advise how to optimize code, i.e. if code is CPU bound, we should focus on ILP optimizations; while if the code is memory bound, we should focus on polyhedral transformations or reduce processor frequency to save power.
  • Associated public software [S19]
  • Extended in journal version [P59], PhD thesis [P58] and DECAN framework [P29]
[CK bib / cM bib] [View doc from CK / cM ]
[P64] 2000 Technical report Jaume Abella, Grigori Fursin, Antonio Gonzalez, Joseph Llosa, Michael O'Boyle, Abhishek Prabhat, Olivier Temam, Sid Touati, Xavier Vera and Gregory Watts.
Advanced Performance Analysis.
MHAOTEU ESPRIT project No 24942 technical report M3.D2, 2001
  • Part of EU FP5 MHAOTEU project [J17]
  • Associated public software [S19]
[CK bib / cM bib]
[P65] 2000 Conference Jaume Abella, Sid Touati, A Anderson, C Ciuraneta, J M Codina M Dai, Christine Eisenbeis, Grigori Fursin, Antonio Gonzalez, Joseph Llosa, Michael O'Boyle, A Randrianatoavina, J Sanchez, Olivier Temam, Xavier Vera and Gregory Watts.
The MHAOTEU Toolset for Memory Hierarchy Management.
16th IMACS World Congress on Scientific Computation, Applied Mathematics and Simulation, Lausanne, Switzerland, August 2000
  • Part of EU FP5 MHAOTEU project [J17]
  • Associated public software [S19]
[CK bib / cM bib] [View doc from CK / cM ]
[P66] 2000 Technical report Jaume Abella, Cedric Bastoul, Jean-Luc Bechennec, Nathalie Drach, Christine Eisenbeis, Paul Feautrier, Björn Franke, Grigori Fursin, Antonio Gonzalez, Toru Kisku, Peter Knijnenburg, Joseph Llosa, Michael O'Boyle, Julien Sebot and Xavier Vera.
Guided Transformations.
MHAOTEU ESPRIT project No 24942 technical report M3.D2, 2001
  • Part of EU FP5 MHAOTEU project [J17]
  • Associated public software [S19]
[CK bib / cM bib]
[P67] 1999 National MS thesis Grigori Fursin.
Unifying remote access to high-performance computing systems (HPC) as a web service.
MS Thesis, Moscow Institute of Physics and Technology and Institute of High-Performance Computing Systems of Russian Academy of Sciences, Moscow, Russia, May 1999
  • Concept [M9]
  • MS degree [Z7]
  • Associated public software [S20]
[CK bib / cM bib]
[P68] 1997 National Conference Grigori Fursin.
Measurement of characteristics of neural elements with the aid of personal computer.
Proceedings of the national conference on physical processes in devices of electronic and laser engineering at Moscow Institute of Physics and Technology, Moscow, Russia, 1997
  • Concept [M10, M9]
  • Associated public software [S21, S20]
  • Associated hardware [H1]
[CK bib / cM bib]
[P69] 1995 National Conference Grigori Fursin.
Modeling of processes of learning and recognition in modified neural network.
Proceedings of the national conference on physical processes in devices of electronic and laser engineering at Moscow Institute of Physics and Technology, Moscow, Russia, 1995
[CK bib / cM bib]
[P70] 1995 National Conference Grigori Fursin.
Restoration of symbols with noise by neural network.
Proceedings of the national conference on physical processes in devices of electronic and laser engineering at Moscow Institute of Physics and Technology, pages 112-117, Moscow, Russia, 1995
[CK bib / cM bib]