Here's results from r4952 compared to r4950 in the prior message....wow!!! You sure you didn't remove something important??? :-) An average 34% improvement on my old desktop and a 23% improvement on the new laptop. More speed plus better decode? I like it!
Mike W9MDB M1, M2 0.793,0.651 Jt9 -w 2 0.784,0.652 Jt9 -w 3 0.564,0.491 Jt9_omp -w 2 0.572,0.478 Jt9_omp -w 3 -----Original Message----- From: Michael Black [mailto:mdblac...@yahoo.com] Sent: Tuesday, February 10, 2015 11:45 AM To: 'Michael Black' Subject: RE: OMP/wisdom testing I've now changed my timing to use timer.out instead...so here are the numbers using that instead....same r4950 as below. And my new code runs until it achieves stability on the average for 10 iterations to 3 significant digits. This code should now be portable as-is to Unix. Easily changed to do *.wav if wanted. Here are the specs on the two CPUs involved. I didn't realize my laptop system has TurboBoost. M1 - Intel Xeon X5450@3Ghz Dual CPU -- FSB=1333MHz, L2=12MB, Hyper-Thread=No, Turbo Boost=No M2 - Intel Core i7-4702MQ@2.2Ghz -- DMI2 5GT/s, Smart cache=6MB, Hyper-Thread=No , Turbo Boost=Yes(3.2Ghz) M1, M2 1.15,0.832 Jt9 -w 2 1.17,0.830 Jt9 -w 3 0.896,0.635 Jt9_omp -w 2 0.888,0.635 Jt9_omp -w 3 Mike W9MDB ------------------------------------------------------------------------------ Dive into the World of Parallel Programming. The Go Parallel Website, sponsored by Intel and developed in partnership with Slashdot Media, is your hub for all things parallel software development, from weekly thought leadership blogs to news, videos, case studies, tutorials and more. Take a look and join the conversation now. http://goparallel.sourceforge.net/ _______________________________________________ wsjt-devel mailing list wsjt-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/wsjt-devel
