|
|
| Author |
Thread Statistics | Show CCP posts - 0 post(s) |
Kazuo Ishiguro
House of Marbles
   |
Posted - 2010.06.10 20:56:00 -
[1]
 Originally by: Torque Daisy
Edited by: Torque Daisy on 10/06/2010 16:31:24
the ati client for f@h will suck up an entire cpu core in addition (and is worse than Nvidia) :@ so i do Rosetta instead, leaving gpu free to crunch milkyway which is ballistic on ATI cards, plus a model of the galaxy is extremely cool.
if only folding had a worthwhile ati client, as it is, waste of electricity :(
In my experience (folding on a 5850 @ 3.7k ppd + E6750) it only uses about a quarter of a core, but that will depend on your processor and graphics driver - make sure you have the most recent driver, as it can make quite a difference. For most desktops, it's still much more efficient (in terms of ppd / watt) than folding with the CPU clients.
---
34.4:1 mineral compression |
Kazuo Ishiguro
House of Marbles
|
Posted - 2010.06.10 20:59:00 -
[2]
Originally by: defiler
Running the SMP right now, seems to be on par with multiple clients. I think I'll go back to the latter method, not being able to throttle it down is a bit of a bummer, I don't want to run it at 100% when I'm actually using the computer (heat -> noise).
This client only really shines with rather expensive multicore i5 / i7 processors. If you do go down that route, a dedicated Xeon box can actually outperform the GPU client on most measures.
---
34.4:1 mineral compression |
Kazuo Ishiguro
House of Marbles
|
Posted - 2010.06.12 09:25:00 -
[3]
Originally by: Hieronymus Alexandre
If it's for Science, it must be done!
Er, not always.
---
34.4:1 mineral compression |
Kazuo Ishiguro
House of Marbles
|
Posted - 2010.06.15 21:34:00 -
[4]
Originally by: Epicbeardman
Mine has been sitting at Working 150/15000 for a good half an hour now. Is that right? I'm all for helping the cause, not so much spending my entire evening troubleshooting.
What sort of client is it, and what hardware are you running it on?
Originally by: Benny Hill
Edited by: Benny Hill on 15/06/2010 14:52:29
Have they become able to grasp the ATI graphics cards and create a stable client for them yet? It was always a frustrating endeavor to run their client.
Although the latest GPU client does not perform as well on ATI cards as it does on Nvidia cards, it is fairly stable for most WUs.
---
34.4:1 mineral compression |
Kazuo Ishiguro
House of Marbles
|
Posted - 2010.06.20 11:40:00 -
[5]
Originally by: defiler
I'll keep running it and see if the logs show anything. (see what I did there?)
The logs show you're getting over 16k PPD:
http://kakaostats.com/t.php?t=75407
I presume you're folding on a GPU as well as the 960?
---
34.4:1 mineral compression |
Kazuo Ishiguro
House of Marbles
|
Posted - 2010.07.01 21:39:00 -
[6]
There is currently a problem with one of the F@H assignment servers that is preventing work packets from being distributed to ATI gpus - I expect it'll be fixed fairly soon.
---
34.4:1 mineral compression |
Kazuo Ishiguro
House of Marbles
|
Posted - 2010.07.24 23:24:00 -
[7]
Is d_fly32 a friend of yours?
---
34.4:1 mineral compression |
Kazuo Ishiguro
House of Marbles
|
Posted - 2010.07.25 08:51:00 -
[8]
Originally by: Verone
Started folding on all 6 CPU cores and 2 GPU Cores 
What combination of clients are you using?
---
34.4:1 mineral compression |
Kazuo Ishiguro
House of Marbles
|
Posted - 2010.08.22 11:15:00 -
[9]
Originally by: Derik Wildstar
Pulled in several for the cause not all decided to stick with the program but that's the norm unfortunately... feel like if we can get a large enough low impact group going we would be better off than pulling in some high ppd ringers for a few days/weeks.
It's better than having people take part who have no idea of the main disadvantages - i.e. increased power bills, heat, noise, CPU/GPU wear, and some degree of GPU hogging (the GPU client is supposed to yield to other apps, but this process doesn't always work too well). Folding has costs, just like everything else, and if people consider them before plunging in, there's less potential for tears further down the line.
All contributions are welcome, even if people eventually decide it's not for them.
---
34.4:1 mineral compression |
|
|
| |
|