| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Support
Forum: GPU Support Forum
Thread: Not using my GPU |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 53
|
|
| Author |
|
|
Hot Mess
Cruncher Joined: Jul 1, 2013 Post Count: 28 Status: Offline |
ok
----------------------------------------I think the problem,WCG does not have enough memory for such a project (hence the 7 days period at MCM). One reason why no GPU projects run here?   |
||
|
|
branjo
Master Cruncher Slovakia Joined: Jun 29, 2012 Post Count: 1892 Status: Offline Project Badges: 
|
As it has been stated several times on this forum: WCG doesn't take decision if some sub-project (e.g. MCM, FAAH, CEP2,...) will run application ported for GPU or not. It is in sole decision of research teams running each sub-project.
----------------------------------------It has nothing to do with "lack of storage space" you think WCG has. If there is some "lack of storage space", it is on researchers' machines.  Crunching@Home since January 13 2000. Shrubbing@Home since January 5 2006  [Edit 1 times, last edit by branjo at Jun 25, 2014 8:16:33 PM] |
||
|
|
JmBoullier
Former Community Advisor Normandy - France Joined: Jan 26, 2007 Post Count: 3716 Status: Offline Project Badges:
|
I believe the scientists know the best what pace of progress is optimal for them. They are learning during this process and incorporate such learning into it (see the latest paper from FAAH scientists) and making adjustments/changes for future analyses. For readers who would not be convinced by branjo's opinion this thread To be able to do this, there must be somebody who will proceed the data we are sending back and who will find relevant insights from them. Now, imagine the situation you are suggesting. We crungers would already sent back to scientists all results, but they would proceeded in their own pace. Suddenly they would discover what they discovered now ad would realize that the rest of results (7 giga combinations) are useless and have to be crunched again. I would be disappointed because my (our) effort (and money) would went to trash bin, wouldn't you? Let's do our part of this work and leave the decisions on more proper part of this work  Better tools for AIDS drug research ...eam publishes a paper perfectly illustrates his reasoning. |
||
|
|
branjo
Master Cruncher Slovakia Joined: Jun 29, 2012 Post Count: 1892 Status: Offline Project Badges:
|
Thanks JmBoullier for pointing to that thread
---------------------------------------- Cheers  Crunching@Home since January 13 2000. Shrubbing@Home since January 5 2006 |
||
|
|
bart567
Cruncher Australia Joined: Nov 25, 2012 Post Count: 3 Status: Offline Project Badges: 
|
Feeling the same here - would love a GPU activated project soon! Does anyone know about the NVIDIA CUDA technology? Is that automatically integrated into CPU performance or can we force it on to make BOINC crunch faster?
Thanks :) |
||
|
|
noderaser
Senior Cruncher United States Joined: Jun 6, 2006 Post Count: 297 Status: Offline Project Badges: 
|
CUDA requires the compiling of special apps, which are utilized by the BOINC projects that make use of nVidia cards. All you have to do is make sure you have the correct driver, and options set within BOINC to allow GPU computing.
----------------------------------------The rival GPGPU API is ATI's CAL, though it seems like many projects that support GPUs have gone with OpenCL (standard that works on both) and abandoned CAL for the most part, though not so much with CUDA. I'm not sure if that is because CAL was especially bad, OpenCL implementation on nVidia chips is not as great, or some combination of both. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Sad to hear that the GPU is not used in any of the projects. This is several hundred times faster. Why not spend a little time writing software for the GPU, if it will pay off quickly?
With the advent of software for cracking hashes on the GPU process now takes few minutes. |
||
|
|
Falconet
Master Cruncher Portugal Joined: Mar 9, 2009 Post Count: 3315 Status: Offline Project Badges: 
|
Sad to hear that the GPU is not used in any of the projects. This is several hundred times faster. Why not spend a little time writing software for the GPU, if it will pay off quickly? With the advent of software for cracking hashes on the GPU process now takes few minutes. Because not all software is suitable for GPU calculations. For example, It may not work at all, or it may work just as fast as a CPU core works or it could be even slower than a CPU. Also, not all researchers have the money/time to develop GPU applications.  - AMD Ryzen 5 1600AF 6C/12T 3.2 GHz - 85W - AMD Ryzen 5 2500U 4C/8T 2.0 GHz - 28W - AMD Ryzen 7 7730U 8C/16T 3.0 GHz |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
maybe you're right. But some project running about 10 years - GPU usage would much reduce this time. It's quite a lot of time)
|
||
|
|
Falconet
Master Cruncher Portugal Joined: Mar 9, 2009 Post Count: 3315 Status: Offline Project Badges:
|
Yes but the Fightaids@home project has been running for 10 years because it is open-ended, the scientists keep adding new work. For all we know, it could end tomorrow or in another 10 years.
----------------------------------------And for that project, the scientists have stated that the use of GPU's is not for their application as it wouldn't work (If I remember, important functions of the application simply won't work in GPU's). - AMD Ryzen 5 1600AF 6C/12T 3.2 GHz - 85W - AMD Ryzen 5 2500U 4C/8T 2.0 GHz - 28W - AMD Ryzen 7 7730U 8C/16T 3.0 GHz |
||
|
|
|