F@H - Wine appears to be faster than native Linux cli

[mgpaulus]

This is not definitive, but.....

I found 2 of my machines, 1 windows and 1 linux box that were apparently working the same WU:
Protein 227: P632_TZ3_NAT. I copied the folders to one of my dual proc machines, which is set up
with (2) Celeron 400 @ 450 (FSB will only go up to 75). Fired up one of the cores with wine,
and fired up the other core under native linux.

over 8 hours, the wine version was averaging 14:50 per frame, and the native linux client was
averaging 18:20 per frame.

I realize this is not the most scientific of tests, as I can't exactly verify that the WUs are identical.
And, I should probably run 2 linux jobs side by side, and 2 wine jobs side by side, to see how much
impact the OS has on the prime, or CPU0 job.

But, for a quick, off the cuff test, it looks like wine can add ~22% to the tinker cores. Which is nothing to
sneeze at. After all, how much are people spending to try to get 10-15% more out of an overclock.
Here's 20%+ for FREE

 

[cakin]

I think it was Poof who last year or the year before found that Seti ran faster under Wine - but it wasnt that big a difference.

 

[ProviaFan]

That's quite a significant increase in production. No wonder my former 500MHz Celeron was so freaking slow in F@H native under Linux.