Multiplication seems slow since current multiplication needs a full-scan of B from every rows of A. To reduce the fullscan time, we should have a 2d square algorithm.
/Edward On Mon, Nov 10, 2008 at 10:49 PM, Apache Wiki <[EMAIL PROTECTED]> wrote: > Dear Wiki user, > > You have subscribed to a wiki page or wiki category on "Hama Wiki" for change > notification. > > The following page has been changed by udanax: > http://wiki.apache.org/hama/PerformanceEvaluation > > ------------------------------------------------------------------------------ > > ||<bgcolor="#ececec"> Version ||<bgcolor="#ececec"> Operation > ||<bgcolor="#ececec"> Cluster Size ||<bgcolor="#ececec"> Rows > ||<bgcolor="#ececec"> Columns ||<bgcolor="#ececec"> Total Maps > ||<bgcolor="#ececec"> Total Reduces ||<bgcolor="#ececec"> Time (seconds) > ||<bgcolor="#ececec"> Bytes Written|| > ||Trunk 712655 ||Addition ||2 node ||1,000 ||1,000 ||2||2||17 seconds > ||66,326,104|| > + ||Trunk 712655 ||SIMDMult ||2 node ||300 ||300 ||2||2|| x || x || > - ||Trunk 712655 ||Multiplication ||2 node ||200 ||200 ||10||1||41 seconds > ||1,277,170,348|| > + ||Trunk 712655 || SIMDMult ||2 node ||200 ||200 ||10||1||41 seconds > ||1,277,170,348|| > - ||Trunk 712655 ||Multiplication ||2 node ||300 ||300 ||60||2||315 seconds > ||7,070,789,092|| > + ||Trunk 712655 || SIMDMult ||2 node ||300 ||300 ||60||2||315 seconds > ||7,070,789,092|| > + || || || || || || || || || || > + ||Trunk 712658 || SIMDMult ||2 node ||300 ||300 ||2||2||181 seconds || > 5,929,512 || > + ||Trunk 712658 || SIMDMult ||2 node ||200 ||200 ||10||1||72 seconds || > 4,931,106 || > - ||Trunk 712658 ||Multiplication ||2 node ||300 ||300 ||60||2||267 seconds > ||16,507,848|| > + ||Trunk 712658 || SIMDMult ||2 node ||300 ||300 ||60||2||267 seconds > ||16,507,848|| > + > > Dependencies Information : > > -- Best Regards, Edward J. Yoon @ NHN, corp. [EMAIL PROTECTED] http://blog.udanax.org
