https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/Linpack-Benchmark-on-Windows7-and-CentOS/m-p/763987#M109 You didn't mention your KMP_AFFINITY setting. It's more critical on Windows, but may help linux also.<BR />Running with HT enabled may be problematic, even when you set 1 thread per core. Thu, 12 Aug 2010 20:29:29 GMT TimP 2010-08-12T20:29:29Z https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/Linpack-Benchmark-on-Windows7-and-CentOS/m-p/763986#M108 <BR /> I had run Intel Optimized LINPACK Benchmark on Windows 7 and CentOS 5.5.<BR /> Platform is DP server and CPU is X5680 (3.33GHz) and memory is 24GB.<BR /> (Benchmark is downloaded from here)<BR /> <A href="http://software.intel.com/en-us/articles/intel-math-kernel-library-linpack-download/" target="_blank">http://software.intel.com/en-us/articles/intel-math-kernel-library-linpack-download/</A><BR /> <BR /> I had tested this benchmark without any VM and both is standalone OS environment. And BIOS setting is same. <BR /> <BR /> <PRE>Intel LINPACK data<BR /><BR />Current date/time: Thu Aug 12 13:24:25 2010<BR /><BR />CPU frequency: 3.333 GHz<BR />Number of CPUs: 12<BR />Number of threads: 12<BR /><BR /><BR /></PRE>Benchamrk number on Windows 7 is...<BR /> <BR /> <PRE>Size LDA Align. Average Maximal<BR />10000 10000 4 103.7096 104.5793<BR />15000 15000 4 108.3996 111.6880<BR />20000 20000 4 132.7701 133.2039<BR />25000 25000 4 106.5018 106.8729<BR />30000 30000 4 136.6855 136.6855<BR />35000 35000 4 109.1213 109.1213<BR />40000 40000 4 139.4487 139.4487<BR /><BR />Benchmark number on CentOS is...<BR /><BR /><PRE>Size LDA Align. Average Maximal<BR />20000 20016 4 144.3745 144.5568<BR />22000 22008 4 146.2399 146.3934<BR />25000 25000 4 147.2634 147.2977<BR />26000 26000 4 148.0729 148.1387<BR />27000 27000 4 148.4799 148.4799<BR />30000 30000 1 148.9899 148.9899<BR />35000 35000 1 149.3800 149.3800<BR />40000 40000 1 150.9887 150.9887<BR />45000 45000 1 151.9700 151.9700<BR /><BR /><BR />Performance number on Windows 7 is much lower than result on CentOS.<BR />Do I need to check something to run on Winodws 7?<BR /><BR />Thank you in advance..<BR /></PRE></PRE><PRE><BR /><BR /></PRE> Thu, 12 Aug 2010 15:41:38 GMT https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/Linpack-Benchmark-on-Windows7-and-CentOS/m-p/763986#M108 Tomuro__Takahiko 2010-08-12T15:41:38Z https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/Linpack-Benchmark-on-Windows7-and-CentOS/m-p/763987#M109 You didn't mention your KMP_AFFINITY setting. It's more critical on Windows, but may help linux also.<BR />Running with HT enabled may be problematic, even when you set 1 thread per core. Thu, 12 Aug 2010 20:29:29 GMT https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/Linpack-Benchmark-on-Windows7-and-CentOS/m-p/763987#M109 TimP 2010-08-12T20:29:29Z