Re: Mathematica 7.0 slow on OS X
- To: mathgroup at smc.vnet.net
- Subject: [mg93812] Re: Mathematica 7.0 slow on OS X
- From: Bill Rowe <readnews at sbcglobal.net>
- Date: Wed, 26 Nov 2008 05:15:05 -0500 (EST)
On 11/25/08 at 7:18 AM, maderri2 at gmail.com (magma) wrote: >Does the Mathematica 7.0 own benchmarkreport take advantage of >parallel computing? There should be an improvement at least there. >Or not? Yes. That is looking at the code in Benchmark.m, there are points where the code is set to use parallel processing when possible. But this is also true for the code in Version 6. At the momement, I don't have a setup that allows me to see what effect this has. When I run version 6 and version 7 of the built in benchmark, I get insignificant differences for all of the tests except 5 (elementary functions), 12 (polynomial expansion) and 15 (solving a linear system). My criteria for significance was the difference in timing between version 6 and 7 was larger than the difference between two subsequent runs of version 6. For polynomial expansion, the code in version 6 and 7 appears identical. Version 7 takes ~30% (3.3 sec) longer on my machine. For the other two tests, the code is not the same. In version 6 of the code there is a function (RandomArray) defined for generating random values. In version 7 of the code, random values are generated by calling RandomReal etc directly. For test 5, version 6 of took ~33% (11.1 sec) longer while for test 15 version 7 took ~5% (1.2 sec) longer. It is tempting to ascribe the differences in these last tests to the difference in code. But looking at the definition of RandomArray it is apparent that it simply calls RandomReal etc. So, I would expect this to add a very small overhead to the timing that would be much smaller than the differences reported. Likely the difference in the last test isn't significant. Had I run version 6 more times, I suspect I would have seen more variation and the difference between version 7 and 6 would be within variation seen in version 6 numbers.