SSS is an ongoing project on high-performance MapReduce implementation in AIST. SSS uses distributed Key Value Store as a base storage, instead of file system like HDFS. 

Our project goal includes; 
  • Implementing a MapReduce framework called 'SSS'.
  • Implementing a special purpose language for MapReduce.
  • Implementing real world application on the system.

  • SSS-mapreduce 1.0 is released.
    We are pleased to announce the 1.0 version of the SSS-mapreduce is now available at https://code.google.com/p/sss-mapreduce/ . Feel free to download and test it. 
    Let us know your findings with it.
    Posted Feb 27, 2013, 11:46 PM by Hidemoto Nakada
  • IPSJ SIG-HPC @ Awara-onsen
    We had a domestic IPSJ SIG-HPC workshop at Awara-onsen, Fukui prefecture. 
    We made presentation there focusing on Graph processing using the merge-reducer of SSS-mapreduce, titled 'SSS-MapReduceのグラフ処理アプリケーションによる評価'.

    Fukui-prefecture is famous for fossiles of dinosaurs, BTW.
    Posted Feb 27, 2013, 11:41 PM by Hidemoto Nakada
  • eScience 2012 poster
    We made poster presentation at eScience 2012, held at Chicago, Oct.9-12/2012, titled 'Stream Processing with Bigdata by SSS-MapReduce'.
    Posted Oct 16, 2012, 1:13 AM by Hidemoto Nakada
  • Two Domestic workshop papers
    10/4/2012, we presented a a paper at SIG-HPC, IPSJ, titled
        MapReduce処理系SSSにおけるContinuous MapReduceの実装 (Continuous MapReduce implementation with SSS-MapReduce)

    10/12/2012, we presented another paper at SIG-CPSY, IEICE, titled
    PrefixSpan法のMapReduce実装の改良 (Improvement of MapReduce implementation of PrefixSpan Method)
    Posted Oct 16, 2012, 1:10 AM by Hidemoto Nakada
  • SWoPP Tottori(鳥取)
    We made following presentations at SWoPP Tottori, a Japanese domestic workshop for parallel processing.

    • MapReduce処理系SSSにおけるKey Value Store アクセス手法の改良
    • MapReduce処理系SSSのアプリケーションによる消費電力測定

    Tottori is famous for it's (rather small) dune. 

    Posted Aug 9, 2012, 6:54 PM by Hidemoto Nakada
Showing posts 1 - 5 of 13. View more »