The .33 inverse for 3 GPU work units, on cards that arn't high end use .5 for 2 GPU work units. My BRP4G took 1hour 34 1/2 minutes doing all 3 wu. The R9 290 has weaker floating point but many more cores.
Our application should pick up all the compatible GPUs present in the system automatically. However, at the moment we do not support multi-GPU configuration options.
I have noticed that the pc that has an Nvidia gpu does not seem to do work other than cpu jobs...I have never had Any probs with Nvidia gpus....I just notice that the pcs with AMD/Ati cards seem to do More work than the Nvidia unit.
Perhaps under "Computing Preference" you do not have "Use GPU while computer is in use".
The reason I say this is that I am "playing" at the present time with a GTX 460 and it is way slow but it is processing for the past couple of weeks when you say you have not had any GPU Nvidia activity CE 3 1.5 cuda.
your computer has a GeForce 9400 GT card, which is not supported by the current applications we run. In the future this may change, but at the moment your GPU will not be contributing.
Hey,
I have some technical questions.
I'd like to know which gpus are better for ce projects? Also are those gpu projects all written in opencl or are the nvidia ones written in cuda?
Apps on CE come and go. At present both AMD and Nvidia cards are best. (The current Nvidia app uses CUDA.)
Open CL benchmarks if anyone's interested...
http://compubench.com/compare.jsp?benchmark=clb11
How do you set up your GPU to run simultanious instances of Einstein? Mine fluxuates wildly almost no load to 87% under AMD Catalyst Control Center.
file name is app_config.xml
location is a hidden folder Program Data, Boinc, Projects, einstein.phys.uwm.edu
[code]
<app_config>
<app>
<name>einsteinbinary_BRP4G</name>
<gpu_versions>
<gpu_usage>.33</gpu_usage>
<cpu_usage>.34</cpu_usage>
</gpu_versions>
</app>
<app>
<name>einstein_S6CasA</name>
<gpu_versions>
<gpu_usage>.33</gpu_usage>
<cpu_usage>.34</cpu_usage>
</gpu_versions>
</app>
[/code]
The .33 inverse for 3 GPU work units, on cards that arn't high end use .5 for 2 GPU work units. My BRP4G took 1hour 34 1/2 minutes doing all 3 wu. The R9 290 has weaker floating point but many more cores.
Now what about the settings for Charity Engine?
-
Our application should pick up all the compatible GPUs present in the system automatically. However, at the moment we do not support multi-GPU configuration options.
I have noticed that the pc that has an Nvidia gpu does not seem to do work other than cpu jobs...I have never had Any probs with Nvidia gpus....I just notice that the pcs with AMD/Ati cards seem to do More work than the Nvidia unit.
Ed
Perhaps under "Computing Preference" you do not have "Use GPU while computer is in use".
The reason I say this is that I am "playing" at the present time with a GTX 460 and it is way slow but it is processing for the past couple of weeks when you say you have not had any GPU Nvidia activity CE 3 1.5 cuda.
DocH,
your computer has a GeForce 9400 GT card, which is not supported by the current applications we run. In the future this may change, but at the moment your GPU will not be contributing.
Thx for the info, Rytis!
One of the PCs I have has an A6-3620 APU with an ATI Raeon HD6530 D GPU, and it seems to do a lot of work so at least both Pcs can contribute to CE!
Ed