Estimated ppd/ rig

[jdsparks21]

Anyone have a run down of what certain rigs will put out as far as ppd's?

I've seen the gpu one but haven't seen a cpu one yet.

I have a Gateway laptop with a e5300 in it 4gb ram and was wondering what it would put out and if it's worth it to let Gixxerguy fold on it. Which client would be best and how to configure it?

Chime in guys.

Also my sig rig: How many ppd's will the gts250 put out? stock compared to overclocking it.

[harlam357]

Hi jd... :welcome: to Team 32! :beer: New guy stocks the fridge. :)

I suppose you're helping out GIX right? I'm not a fan of laptop folding... a lot of folks do it and don't have any issues. I'm not going to advise you one way or the other... and I doubt anyone will have any readily available data on that particular chip.

So is there any place where this type of production data exists? No, not yet... I'm hopefully going to put together something as such, using HFM as a data conduit... it just hasn't come together yet.

The GTS250 will probably make between 4000~6000ppd on current WUs... but that's just a guess, I don't have any G92 based gpus running anymore so I don't know how they are fairing... I read they're doing real nice on the p3469 and p3470 varieties however.

[ChasR]

The Win SMP2 client should turn 13:00/frame on a p6024, 1950 ppd. The GTX250 running shaders at 1998 makes about 5800 ppd on the worst current WUs and 8900 ppd on the best. The average is about 6400 ppd. I haven't run one at stock in so long, I don't know what it'll do. I did post my testing on Overclocking on a G92 chip over a year ago, that might have mentioned the effect of core, shader, and memory clocking individually and compared back to stock. I'd probably get quite different results on some of the current WUs.

[harlam357]

...I doubt anyone will have any readily available data on that particular chip.

Let me rephrase that... if ChasR doesn't have any data on that chip, I doubt anyone else will. :D

[Mark620]

oops!
Harlem opened his mouth and stuck his foldingfoot in it.

[ChasR]

Mine's really an educated guess, no direct data to back it up. I estimated it like this:

An e5300 runs at 2.6 GHz and my T7200 runs at 2.0 GHz. The T7200 runs 16:33/frame on p6024. 16.55/2.6 x 2 = 12.73/frame. I added a little because of the smaller cache on the e5300 and, voila, 13:00/frame. That should be pretty close. I'll check against an e6600 @ 2.6 tomorrow, if I can remember to do it.

[ChasR]

Back in May 09, I found that a three step shader OC (~9%) resulted in a 7% increase in ppd on a G92.