https://community.intel.com/t5/Intel-MPI-Library/Intel-optimised-Linpack-scaling-comparision/m-p/954323#M3084 <DIV></DIV> <P>Bala -</P> <P>At only 1000 rows and columns, the workload is likely too small to sustain good speedup for 16 threads (just over 60 rows per thread). System and threading overheads are likely taking a relatively larger fraction of time to the work being done, which reduces Gflops.</P> <P>What kind of speeds do you get for larger values of n (e.g., 5000, 10000, 20000)?</P> <P>--clay</P> <DIV></DIV> Wed, 04 Jan 2006 01:34:36 GMT ClayB 2006-01-04T01:34:36Z https://community.intel.com/t5/Intel-MPI-Library/Intel-optimised-Linpack-scaling-comparision/m-p/954322#M3083 The intel optimised linpack (n=1000) that is run a 16 way server has given the following results that is causing a little concern.<BR /><BR />with<BR />1 thread and 1 cpu : 5.8731 Gflops<BR />4 thread and 4 cpu : 15.101 Gflops<BR />8 thread and 8 cpu : 15.107 Gflops<BR />16 thread and 16 cpu : 12.471 Gflops<BR /><BR />I am not to trace what could be the problem?<BR />Please help.<BR /><BR />Bala Mon, 02 Jan 2006 16:25:30 GMT https://community.intel.com/t5/Intel-MPI-Library/Intel-optimised-Linpack-scaling-comparision/m-p/954322#M3083 Balamurugan_R_ 2006-01-02T16:25:30Z https://community.intel.com/t5/Intel-MPI-Library/Intel-optimised-Linpack-scaling-comparision/m-p/954323#M3084 <DIV></DIV> <P>Bala -</P> <P>At only 1000 rows and columns, the workload is likely too small to sustain good speedup for 16 threads (just over 60 rows per thread). System and threading overheads are likely taking a relatively larger fraction of time to the work being done, which reduces Gflops.</P> <P>What kind of speeds do you get for larger values of n (e.g., 5000, 10000, 20000)?</P> <P>--clay</P> <DIV></DIV> Wed, 04 Jan 2006 01:34:36 GMT https://community.intel.com/t5/Intel-MPI-Library/Intel-optimised-Linpack-scaling-comparision/m-p/954323#M3084 ClayB 2006-01-04T01:34:36Z