tag:support.primatelabs.com,2011-01-31:/discussions/geekbench/563-64bit-performance-on-linux-and-windowsPrimate Labs: Discussion 2012-07-20T12:23:49Ztag:support.primatelabs.com,2011-01-31:Comment/163664692012-06-02T02:04:51Z2012-06-02T02:04:51Z64bit performance on linux and windows<div><p>Hi Nasos,</p>
<p>Thanks for your message. Geekbench 2 tries to measure the
performance users can expect from their applications, rather than
the peak performance their hardware provides. This means using the
de-factor standard compiler on each platform (GCC on Mac OS X and
Linux, Visual C++ on Windows) with the vendor-recommended
optimization flags.</p>
<p>This approach does result in different scores on different
platforms. It's often the case, though, that the same application
will perform differently on different platforms.</p>
<p>If you're interested in only comparing the relative performance
of the underlying hardware instead, I recommend performing all of
your testing on one operating system. That way you'll minimize the
number of variables that change (e.g., operating system, compiler,
etc) from system to system.</p>
<p>Best,<br>
John</p></div>Johntag:support.primatelabs.com,2011-01-31:Comment/163664692012-07-10T20:20:09Z2012-07-10T20:20:10Z64bit performance on linux and windows<div><p>Can one make a meaningful comparison between Android, iOS, Mac,
and Windows? Are they doing the exact same amount of work like
loops, indices, etc?</p></div>Bjorntag:support.primatelabs.com,2011-01-31:Comment/163664692012-07-10T20:41:05Z2012-07-10T20:41:05Z64bit performance on linux and windows<div><p>Bjorn -</p>
<p>Geekbench 2 uses the same source code on all platforms (Android,
iOS, Linux, Mac, and Windows) meaning you can compare results on
all five platforms.</p>
<p>Best,<br>
John</p></div>John