Does Geekbench version impact on results (32 bit tryout verses 64bit) ?
I recently tried Geekbench tryout (32 bit test) here, score of 11582:
http://browser.primatelabs.com/geekbench2/1570405
Looked for a close comparison (exact same hardware and OS level etc) and found this record, score of 16194:
http://browser.primatelabs.com/geekbench2/1540854
I'm curious as to the large difference - the only thing I can see that is different in the comparison (second link above) is that they are using the 64 bit version of Geekbench; would that impact so much on the benchmark outcome ? Seems to be big differences in the floating point performance in particular.
Keyboard shortcuts
Generic
| ? | Show this help |
|---|---|
| ESC | Blurs the current field |
Comment Form
| r | Focus the comment reply box |
|---|---|
| ^ + ↩ | Submit the comment |
You can use Command ⌘ instead of Control ^ on Mac
1 Posted by FORTINUEX on 27 Jan, 2013 01:38 AM
The reason the big jump in Floating point in 64 bit test is the lack of restriction from 32 bit processing when you have a multithreaded processor, seems self explanatory but that is probably the simplest answer. I can tell you personally RAM speed and Stream performance, though there is a slight point increase maybe 500-1000 on overall score but this improves your final score a lot more then some think. If you are running a 64 bit system on a 64 bit Operating System you have to take the 32 bit tests with a grain of salt per say. Blur Image multi-core scalar thrives on cores and threads in 64 bits with the extra "pipelines". I personally think the other Multi-core tests need tweaked and the barely noticeable increase doesn't help with proving 64 bit is much better.