ILP-Based Resource-Aware Compilation

Palsberg, Jens; Naik, Mayur

doi:10.1016/b978-012385251-9/50027-x

Cited by 6 publications

(3 citation statements)

References 44 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Notably, uses of ILP in register allocation, code-generation, and optimization ordering for conventional architectures [32,42] are unrelated to the primitives of spatial architecture scheduling. Affine loop analysis and resulting instruction scheduling/code-generation for superscalar processors is a popular use of mathematical models [2,5,44], and since it falls within the data-dependence analysis role of the compiler, not its scheduler, is a non-goal for us.…”

Section: Non Goalsmentioning

confidence: 99%

A general constraint-centric scheduling framework for spatial architectures

et al. 2013

View full text Add to dashboard Cite

Specialized execution using spatial architectures provides energy efficient computation, but requires effective algorithms for spatially scheduling the computation. Generally, this has been solved with architecture-specific heuristics, an approach which suffers from poor compiler/architect productivity, lack of insight on optimality, and inhibits migration of techniques between architectures. Our goal is to develop a scheduling framework usable for all spatial architectures. To this end, we expresses spatial scheduling as a constraint satisfaction problem using Integer Linear Programming (ILP). We observe that architecture primitives and scheduler responsibilities can be related through five abstractions: placement of computation, routing of data, managing event timing, managing resource utilization, and forming the optimization objectives. We encode these responsibilities as 20 general ILP constraints, which are used to create schedulers for the disparate TRIPS, DySER, and PLUG architectures. Our results show that a general declarative approach using ILP is implementable, practical, and typically matches or outperforms specialized schedulers.

show abstract

Section: Non Goalsmentioning

confidence: 99%

A general constraint-centric scheduling framework for spatial architectures

et al. 2013

View full text Add to dashboard Cite

show abstract

Section: Non Goalsmentioning

confidence: 99%

A general constraint-centric scheduling framework for spatial architectures

Nowatzki

Sartin-Tarm

Carli

et al. 2013

Proceedings of the 34th ACM SIGPLAN Conference on Programming Language Design and Implementation

View full text Add to dashboard Cite

Specialized execution using spatial architectures provides energy efficient computation, but requires effective algorithms for spatially scheduling the computation. Generally, this has been solved with architecture-specific heuristics, an approach which suffers from poor compiler/architect productivity, lack of insight on optimality, and inhibits migration of techniques between architectures.Our goal is to develop a scheduling framework usable for all spatial architectures. To this end, we expresses spatial scheduling as a constraint satisfaction problem using Integer Linear Programming (ILP). We observe that architecture primitives and scheduler responsibilities can be related through five abstractions: placement of computation, routing of data, managing event timing, managing resource utilization, and forming the optimization objectives. We encode these responsibilities as 20 general ILP constraints, which are used to create schedulers for the disparate TRIPS, DySER, and PLUG architectures. Our results show that a general declarative approach using ILP is implementable, practical, and typically matches or outperforms specialized schedulers.

show abstract

“…Integer Linear programming is widely used to obtain good exact and approximation algorithms for different problems, see i.e. [24,44,72,75,78]. The advantage of this approach is that it is very generic and, therefore, we can apply many different well-developed LP-techniques to obtain better quality solutions or to speed-up the algorithm.…”

Section: Chapter 6 Integer Linear Programming Formulations For Treewidthmentioning

confidence: 99%

Exploiting geometric properties in combinatorial optimization = Benutten van geomwetrische eigenschappen in combinatiorische optimalisatie

Usotskaya¹

View full text Add to dashboard Cite

People interested in the research are advised to contact the author for the final version of the publication, or visit the DOI to the publisher's website.• The final author version and the galley proof are versions of the publication after peer review.• The final published version features the final layout of the paper including the volume, issue and page numbers. Link to publication General rightsCopyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.• Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain • You may freely distribute the URL identifying the publication in the public portal.This thesis would never appear without the endless support of some people. I am glad to name them here. First of all, many thanks go to my enormously patient supervisor, Alexander Grigoriev. He was always full of enthusiasm and new ideas about any project.Part of the research was made in close cooperation with André Berger and Bert Marchal. I am very grateful for the knowledge, ideas and precision they brought into the thesis. My special thanks go to Bert for his patience and for always being there for me.The unconditional support of my parents led me from the secondary school to this doctoral thesis. Without their understanding and help I would never proceeded so far. Thank you, mom and dad.Special thanks to my beloved roommate Bas, for charts and the guidance through the Dutch culture and humor. His regular jokes and shiny smiles made rainy days much brighter. He was not the only one who lightened my life during these years. Without mentioning their names I thank all the friends and colleagues from the university, who made my life in Maastricht a very enjoyable and precious time.A lot of thanks go to Karin and Haydeé for their help with all the paper work. These pretty ladies make the atmosphere in the department pleasant and relaxing.I thank Rudolf Müller, Jean Cardinal and Ralf Peeters for refereeing the thesis and for all the useful comments they provided to make the text more clear and readable.The research that forms the basis of this thesis was conducted from 2008 to 2011 at the Quantitative Economics department; it was funded by the department and by Maastricht Research School of Economics of Technology and Organizations (ME-TEOR). This support is gratefully acknowledged.

show abstract

ILP-Based Resource-Aware Compilation

Cited by 6 publications

References 44 publications

A general constraint-centric scheduling framework for spatial architectures

A general constraint-centric scheduling framework for spatial architectures

A general constraint-centric scheduling framework for spatial architectures

Exploiting geometric properties in combinatorial optimization = Benutten van geomwetrische eigenschappen in combinatiorische optimalisatie

Contact Info

Product

Resources

About