ストリーム指向プログラムのマクロ並列化の研究

【 研究概要 】

ストリーム指向プログラムはGPUといったメニーコアアクセラレータの普及によって、科学技術計算から産業用製品にまで利用されている。その単体性能は、チップ内における密並列によるプログラム実行により高い性能を示す。しかし、複数のアクセラレータを使った超並列計算を考慮すると、タスクの分割と通信タイミングを配慮したプログラム開発が必要になり、性能をスケーラブルに維持したままの開発が困難となる。本研究では、このようなGPUでのストリーム指向プログラムを対容積・対電力での計算能力の高密度化をねらい、自動的に複数のGPUで並列化し、スケーラブルに性能向上が可能なプログラミング基盤技術を開発する。Read More

Research articles in Journals and Magazines

  1. Shinichi Yamagiwa, Guyue Wang, Koichi Wada. Development of an Algorithm for Extracting Parallelism and Pipeline Structure from Stream-based Processing flow with Spanning Tree, International Journal of Networking and Computing, 159--179, (2015-01-10), DOI:10.15803/ijnc.5.1_159
  2. Shinichi Yamagiwa, Leonel Sousa. Modelling and programming stream-based distributed computing based on the meta-pipeline approach, International Journal of Parallel, Emergent and Distributed Systems, 311--330, (2009-07-29), DOI:10.1080/17445760902719968
  3. Shinichi Yamagiwa, Leonel Sousa. Caravela: A Novel Stream-Based Distributed Computing Environment, Computer, 70--77, (2007-05-05), DOI:10.1109/mc.2007.161

Conference Proceedings

  1. Shinichi Yamagiwa, Ryo Jozaki, Shixun Zhang, Ryo Zaizen, Dewen Xu. Exploiting Execution Order and Parallelism from Processing Flow Applying Pipeline-Based Programming Method on Manycore Accelerators, 2013 42nd International Conference on Parallel Processing, , (2013-10-01), DOI:10.1109/icpp.2013.84
  2. Shinichi Yamagiwa, Ryoyu Watanabe, Koichi Wada. Operation Synchronization Technique on Pipeline-Based Hardware Synthesis Applying Stream-Based Computing Framework, 2013 IEEE International Symposium on Parallel & Distributed Processing, Workshops and Phd Forum, , (2013-05-20), DOI:10.1109/ipdpsw.2013.61
  3. Shinichi Yamagiwa, Masahiro Arai, Koichi Wada. Efficient handling of stream buffers in GPU stream-based computing platform, Proceedings of 2011 IEEE Pacific Rim Conference on Communications, Computers and Signal Processing, , (2011-08-23), DOI:10.1109/pacrim.2011.6032907
  4. Shinichi Yamagiwa, Leonel Sousa. CaravelaMPI: Message Passing Interface for Parallel GPU-Based Applications, 2009 Eighth International Symposium on Parallel and Distributed Computing, , (2009-06-30), DOI:10.1109/ispdc.2009.24
  5. Shinichi Yamagiwa, Leonel Sousa. Design and implementation of a tool for modeling and programming deadlock free meta-pipeline applications, 2008 IEEE International Symposium on Parallel and Distributed Processing, , (2008-04-14), DOI:10.1109/ipdps.2008.4536121
  6. Shinichi, Yamagiwa, Diogo, Ricardo, Cardoso Antao, Leonel, Sousa. Design and implementation of a graphical user interface for stream-based distributed computing, the IASTED International Conference on Parallel and Distributed Computing and Networks (PDCN 2008), , (2008-02-12), DOI:
  7. Gabriel Falcao, Shinichi Yamagiwa, Vitor Silva, Leonel Sousa. Stream-Based LDPC Decoding on GPUs, First Workshop on General Purpose Processing on Graphics Processing Units (GPGPU), , (2007-10-01), DOI:
  8. Shinichi Yamagiwa, Leonel Sousa, Tomas Brandao. Meta-Pipeline: A New Execution Mechanism for Distributed Pipeline Processing, Sixth International Symposium on Parallel and Distributed Computing (ISPDC2007), , (2007-07-05), DOI:10.1109/ispdc.2007.36
  9. Shinichi Yamagiwa, Leonel Sousa. Design and implementation of a stream-based distributedcomputing platform using graphics processing units, Proceedings of the 4th international conference on Computing frontiers - CF2007, , (2007-05-07), DOI:10.1145/1242531.1242561
  10. Leonel Sousa, Shinichi Yamagiwa. Caravela: A Distributed Stream-Based Computing Platform, 3rd HiPEAC Industrial Workshop, IBM Haifa Labs, Israel, , (2007-04-17), DOI:

主な獲得研究予算