I use ReduceMesh in Holomark2 to test the system on a 370K model, as it is reliable and uses all cores. But today when I was testing the laptop (i5 dual core) in software mode (turned off OpenGL) the reduce mesh took 251 seconds instead of 27! Why is that? Does it use OpenGL, Cuda or OpenCL in any way?
No it doesn’t. However, if you are running in software mode, you are using your processor to generate the display - and this will presumably take cycles away from the decimator.
Have you had a look at your processes and threads for the two cases in Task Manager or Process Explorer?
…and was it any help illustrating Andy’s explanation?
I’ll look into it. It “should” not take cycles at all, as the display is static, so it should not redraw any thing, as far as I can understand.
On my workstation with the Quadro I see no difference between having OpenGL on or off. I’ll double check on the laptop and I’ll see if I can reproduce it.
Sorry, I didn’t see Andy’s reply
@andy I did the same testing today, and now it works as expected. I have no idea why the laptop did as it did yesterday.