| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: FightAIDS@Home
Thread: Autodock & CUDA |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 40
|
|
| Author |
|
|
djibeXX
Cruncher Joined: Dec 3, 2005 Post Count: 9 Status: Offline Project Badges: 
|
Dear all,
I saw this project on sourceforge : http://sourceforge.net/projects/gpuautodock/ As fightaids@home uses Autodock, can't we make a special build for nvidia CUDA video cards ??? It would be much faster. Waiting for your answers |
|||
|
||||
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
Unfortunately a paper showed that it was actually not a very good idea to do this on Autodock.
----------------------------------------Note to self: Think to make a CUDA short FAQ, summarizing all the scientists and WCG replies. HCC, projected for implementation in 2010 CEP2, maybe in a next life, if the phase 2 were big enough HCMD2, not suitable per scientists FAAH, NO not feasible (thus could also apply to HFCC) HPF2, never asked. Check with Rosetta@home on the standings DDDT2, too short to even think about it and issues enough to get it even going on CPU RICE - can't remember but think to have read it would not gain. Along these line, with the relevant discussion links
WCG
----------------------------------------Please help to make the Forums an enjoyable experience for All! [Edit 2 times, last edit by Sekerob at Nov 19, 2009 2:26:09 PM] |
||
|
|
djibeXX
Cruncher Joined: Dec 3, 2005 Post Count: 9 Status: Offline Project Badges:
|
Thanks a lot. Nvidia and ATI will soon be leaders in supercomputers ... or I hope so.
|
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
RICE - can't remember but think to have read it would not gain. Here ya go  http://www.worldcommunitygrid.org/forums/wcg/...ead,26275_offset,0#239244 |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
thanks eso, that was the one and on HPF2 ergo Rosetta based science I found a text over there:
----------------------------------------http://boinc.bakerlab.org/rosetta/forum_thread.php?id=5023 Message 62931 - Posted 14 Aug 2009 16:59:13 UTC I spent some time looking into using CUDA for nvidia gpus but it turns out that a pretty extensive rewrite would be required to really make good use of the gpus and CUDA does not support C++ yet. I even tried to port a couple routines but the memory bandwidth was just too much. It will be a long way till OpenCL is mature and greater acceptance is found and scientists start/will see the benefits from when first designing their project with all the special limitations in mind.
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
robertmiles
Senior Cruncher US Joined: Apr 16, 2008 Post Count: 443 Status: Offline Project Badges: 
|
Nvidia has recently announced that they will soon start providing new compilers that support CUDA use from C++, and a few other computer languages such as OpenCL. However, they have NOT made it clear yet whether this is only for the new GT300 series of Nvidia chips that they have barely started selling, if at all. Also, they have NOT made it adequately clear whether parts of the programs still need to be written directly in CUDA.
----------------------------------------http://www.gpugrid.net/forum_thread.php?id=1406 ATI is well behind in providing compilers for computer languages not specific to ATI chips. Neither is likely to do very much good for algorithms that are mostly memory-limited instead of processor-limited; they may only allow running the algorithm on just one GPU core in the graphics chip even if it has hundreds of GPU cores, and since the graphics card clock rates are generally slower than CPU clock rates, the result will be something that can run on a GPU, but slower than on a CPU core. [Edit 2 times, last edit by robertmiles at Nov 20, 2009 4:43:11 AM] |
||
|
|
freesam
Cruncher Poland Joined: Feb 3, 2007 Post Count: 3 Status: Offline Project Badges:
|
but in a much larger number..
/ / sorry I'm using translate.google ale za to w znacznie większej ilości.. //przepraszam używam translate.google |
||
|
|
robertmiles
Senior Cruncher US Joined: Apr 16, 2008 Post Count: 443 Status: Offline Project Badges:
|
but in a much larger number.. / / sorry I'm using translate.google ale za to w znacznie większej ilości.. //przepraszam używam translate.google In a much larger number on graphics card that have much more memory than running the program on just one GPU core requires, but not on those that don't. |
||
|
|
3A4scLiRhJVcdT2K9q9kQNxzxYJ9
Advanced Cruncher Joined: Nov 16, 2009 Post Count: 72 Status: Offline |
Hello everybody... nice to be alive and crunching :)
----------------------------------------I personally really hope that as new projects will come up in the future... that they will support GPU crunching... (not all, but at least some of them...) That would help me quite a lot with my boinc-stats :) Also, I want to use my PC-stuff in the most effective way possible. If I would not be seriously interested in the research done in the med sector, I probably would have got myself half a dozen of GPU cards instead of my dedicaded server and switched to other Boinc projects. In the beginning, it was my intention to stick with WCG only, but as I noted that my (ATI 4890) GPU got some work done in 45seconds which actually would take my 1 CPU core somewhat around an hour to do, I thought that from the effectiveness point of view of the job being done, it really would make sense to join that project even if it does not belong to the WCG-project. I thought that probably other researchers would be happy to get their work done as well, even if from my personal point of view, this could be less interesting for humanity than the med projects here at WCG. I plan to switch back to "WCG only" as soon as WCG supports a project which will run with my ATI. Let's see when this may happen... because a look at my boinc stats below says it all. Everyone, have an excellent weekend! :) -Chris- [Edit 1 times, last edit by SHOPXXL-COM at Nov 27, 2009 9:55:18 AM] |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
There is one more point which is attainable much quicker - 64 bit computing. A lot of people are moving to 64 bits, some companies like toshiba start selling PCs preinstalled with Windows 7 x64 preinstalled.
Now our Fight aids at home application in World community grid comes only in old 32 bit verrsion. It could be easilly recompiled in 64 bits gaining some speed, possibly up to 25%. As more and more computers with x64 bit are popping up, this would be excelent opportunity to speed up project. I do not have source code of the distributed FAAH application, but people at the institute have it and possibly could recompile it or let me or others to do that. This is realtively simple job that could be done meanwhile we go for CUDA or OpenCL solution. |
|||
|
||||
|
|