[RC5] gpu client

Nathan Klassen klassenn at uregina.ca
Tue Feb 24 20:59:56 EST 2009


I just wanted to chime in here.I recently installed the Beta Windows 32bit
client on my main computer, using an 8800GTS 512.
This is different than the 8800GTS 320/640 due to Nvidia's ridiculous naming
conventions.
I've only been running it for an hour or so but it is averaging 350m keys
per second.
Current peak is 361 million keys per second.
That's not a typo. I can post a screenshot if anyone is interested.

Mr. McCoy is running an 8600GT (Stats on it
http://www.nvidia.com/object/geforce_8600.html)<http://www.nvidia.com/object/geforce_8600.html>
32 Stream Processors, 675mhz Core, 1450mhz Shader, 1000mhz Memory Clock,
128bit memory interface.

My 8800GTS 512 (Stats on the 8800 series:
http://www.nvidia.com/page/geforce_8800.html)<http://www.nvidia.com/page/geforce_8800.html>
128 Stream Processors, 650mhz Core, 1625mhz Shader, 1000mhz Memory Clock,
256bit memory interface.

I would assume the number of stream processors and the Shader clock are the
biggest reason for the difference.

I am using CUDA 4 Pipe 64-THD (Option 6). All but one were within 10M
keys/second of each other.

I'll re-run the benchmark's tomorrow to see which one was the slowest.

I have the same 2D Image lag being described by Mr. McCoy and I'm just
running a lowly 1280x1024 19" LCD, so I don't think your 2 large high
resolutions screens are causing your lag. I'll only be running this while
I'm at work / overnight due to the lag, stats will be posted @
masta_EMS at yahoo.com over the next few days.

Thanks to all that worked so hard on bringing this client to life :)

- Nathan Klassen

On Wed, Feb 18, 2009 at 9:17 AM, Kevin McCoy <kgmccoy at idk-inc.com> wrote:

>  SideWinder wrote:
>
>
> I'm not sure, but I believe the problem is not so much in the client being
> "not nice" in terms of scheduling with the GPU, but that the driver has to
> reload lots of stuff to the card's memory when it switches between graphics
> and the CUDA work.
>
> It might be slow because this machine is running dual monitors at 1680 x
> 1050 x 32bit with a 60 Hz refresh. Whats that work out to? 3360x1050? Eek!
> This is my software development machine (no games, some CAD) and I like to
> have a lot of windows open at a time.
>
>
> At least to me it doesn't look like there is any difference between what is
> being drawn in 2D, the refresh rate just sucks when the CUDA Dnetc is
> running. The Vista Aero UI is another issue though.
>
> I'm running XP - too many issues on Vista that I am not willing to deal
> with. I'll let you early adopters bash MS bugs for me :-)
>
>
> Btw, what kind of keyrate do you get with the card and which core is the
> best for it?
>
> It peaks out at ~80 MKeys/S <http://idk.serveftp.net/cowboy/u1/host.html>.
> Click on the link and look down my herd stats table until you see "Samovar"
> (get it? :-) ) Samovar is pretty much my only RC5-72 machine right now,
> since I concentrate almost all my other cows on OGR.
>
> The RC5-72 *average* is a bit less, since I have to pause the thing to get
> any work accomplished.
>
> I am using the CUDA 2-pipe 64-thd core. It worked out best in the long
> benchmark test I ran. Surprisingly, the short benchmark picked another core
> that wasn't nearly as fast. I think the CUDA client developers need to tweak
> the default-core picker a bit.
>
> Best regards,
>
> Kevin G. McCoy
>
>
> _______________________________________________
> rc5 mailing list
> rc5 at lists.distributed.net
> http://lists.distributed.net/mailman/listinfo/rc5
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.distributed.net/pipermail/rc5/attachments/20090224/6bcafb3e/attachment.html 


More information about the rc5 mailing list