Filters








5 Hits in 2.7 sec

Solving Hard MIPLIB2003 Problems with ParaSCIP on Supercomputers: An Update

Yuji Shinano, Tobias Achterberg, Timo Berthold, Stefan Heinz, Thorsten Koch, Michael Winkler
2014 2014 IEEE International Parallel & Distributed Processing Symposium Workshops  
In 2010, two out of six open MIPLIB2003 instances could be solved by ParaSCIP in more than ten consecutive runs, restarting from checkpointing files.  ...  One challenge of today's software development is how to harness this vast computing power in order to solve really hard mixed-integer programming instances.  ...  Acknowledgment We are thankful to the HRLN II and HRLN III supercomputer staff, especially Bernd Kallies, Hinnerk Stüben and Matthias Läuter who gave us support at any time we needed it.  ... 
doi:10.1109/ipdpsw.2014.174 dblp:conf/ipps/ShinanoABHKW14 fatcat:mqv6wze7arardhkr3fj7epggqu

Solving Open MIP Instances with ParaSCIP on Supercomputers Using up to 80,000 Cores

Yuji Shinano, Tobias Achterberg, Timo Berthold, Stefan Heinz, Thorsten Koch, Michael Winkler
2016 2016 IEEE International Parallel and Distributed Processing Symposium (IPDPS)  
To achieve these results we used an enhanced version of ParaSCIP, setting a new record for the largest scale MIP computation: up to 80,000 cores in parallel on the Titan supercomputer.  ...  In this paper we describe the basic parallelization mechanism of ParaSCIP, improvements of the dynamic load balancing and novel techniques to exploit the power of parallelization for MIP solving.  ...  Acknowledgment We are grateful to the HLRN III supercomputer staff, especially Matthias Läuter and Guido Laubender and to the ISM supercomputer staff in Tokyo, especially Tomonori Hiruta.  ... 
doi:10.1109/ipdps.2016.56 dblp:conf/ipps/ShinanoABHKW16 fatcat:tndwdnlatzavhpumxpsrfs6teq

ParaSCIP: A Parallel Extension of SCIP [chapter]

Yuji Shinano, Tobias Achterberg, Timo Berthold, Stefan Heinz, Thorsten Koch
2011 Competence in High Performance Computing 2010  
We present ParaSCIP, an extension of SCIP, which realizes a parallelization on a distributed memory computing environment.  ...  Using a first implementation of ParaSCIP, we were able to solve two previously unsolved instances from MIPLIB2003, a standard test set library for MIP solvers.  ...  Solving open instances from MIPLIB2003 on HLRN II In this section, we present computational results for solving two open problem instances, ds and stp3d, from MIPLIB2003 which were conducted on the HLRN  ... 
doi:10.1007/978-3-642-24025-6_12 dblp:conf/chpc/ShinanoAB0K10 fatcat:6lgwfilfjrgydm7ptal2oidy6y

An Easy Way to Build Parallel State-of-the-art Combinatorial Optimization Problem Solvers: A Computational Study on Solving Steiner Tree Problems and Mixed Integer Semidefinite Programs by using ug[SCIP-*,*]-Libraries

Yuji Shinano, Daniel Rehfeldt, Tristan Gally
2019 2019 IEEE International Parallel and Distributed Processing Symposium Workshops (IPDPSW)  
Branch-and-bound (B&B) is an algorithmic framework for solving NP-hard combinatorial optimization problems.  ...  Therefore, in order to solve hard problem instances for the first time, one needs to accelerate state-of-the-art algorithm implementations.  ...  Acknowledgment The authors would like to thank Chuen Teck See for first experimenting with a parallel version of SCIP-SDP, and Marc E.  ... 
doi:10.1109/ipdpsw.2019.00095 dblp:conf/ipps/ShinanoRG19 fatcat:k3d3t4rhsvfu3d3m7yfft2d5d4

FiberSCIP—A Shared Memory Parallelization of SCIP

Yuji Shinano, Stefan Heinz, Stefan Vigerske, Michael Winkler
2018 INFORMS journal on computing  
There is a parallel extension of SCIP named ParaSCIP, which parallelizes SCIP on massively parallel distributed memory computing environments.  ...  This paper focuses on a parallelization of SCIP (Solving Constraint Integer Programs), which is a MIP solver and constraint integer programming framework available in source code.  ...  Since ParaSCIP targets on solving very hard instances, where the overhead could be neglected, FiberSCIP is intended to run on a normal desktop computer with multi-core processor(s).  ... 
doi:10.1287/ijoc.2017.0762 fatcat:kmwwzfzxrvbwzpt4t5kgawvwza