| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Official Messages
Forum: News
Thread: May Update: OpenPandemics - COVID-19 |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 10
|
|
| Author |
|
|
WCGAdmin
World Community Grid Admin Joined: Jun 9, 2020 Post Count: 161 Status: Offline |
The project has added GPU power to the existing strong CPU power that supports research for potential COVID-19 treatments.
https://www.worldcommunitygrid.org/about_us/viewNewsArticle.do?articleId=705 |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
... 10 months CPU work done in 8 GPU days. Seriously considering to move my 12 OPN1 CPU threads onto a project that does not have GPU capability.
(of course the question is, and the return to 1700 GPU tasks per 30 minutes seems to infer, do the scientists even have the capacity to analyze the data at 37.5 times the old speed.) |
||
|
|
Sgt.Joe
Ace Cruncher USA Joined: Jul 4, 2006 Post Count: 7223 Status: Offline Project Badges: 
|
(of course the question is, and the return to 1700 GPU tasks per 30 minutes seems to infer, do the scientists even have the capacity to analyze the data at 37.5 times the old speed.) I will be speculating there has to be some automated procedures to analyze the returned results, whether they are generated from cpu or gpu. Probably something along the lines of looking at the binding energy numbers and only further analyzing those who meet some predetermined threshold. After all, we are doing the preliminary screening to help narrow down the particular molecules or compounds for further testing. If we can get the millions of possibilities down to 50-100 that will hopefully yield some potential drug that can clear a sick person's system from this scourge. Crunch on. Cheers
Sgt. Joe
*Minnesota Crunchers* |
||
|
|
yoerik
Senior Cruncher Canada Joined: Mar 24, 2020 Post Count: 413 Status: Offline Project Badges: 
|
... 10 months CPU work done in 8 GPU days. Seriously considering to move my 12 OPN1 CPU threads onto a project that does not have GPU capability. (of course the question is, and the return to 1700 GPU tasks per 30 minutes seems to infer, do the scientists even have the capacity to analyze the data at 37.5 times the old speed.) yeah... same. For now I am reducing the number of OPN1 WUs, and will turn it off altogether once the project can resume sending out work to GPUs. I'm glad I was able to contribute this past year, but the GPU test has made it abundantly clear that the project is best served if it puts its time into supplying GPU WUs. 10 months of work in 8 days --- they can still send CPU WUs if they want; but GPU deserves the vast majority of the WUs in my book.  |
||
|
|
DE KOMETEN
Cruncher Joined: Apr 27, 2007 Post Count: 24 Status: Offline Project Badges: 
|
Am thinking too about diminishing my daily CPU work, since GPU's will generate so much more results.
----------------------------------------The only thing now is that the distribution of GPU WU's is something like a lottery. I havent got any GPU work since the beginning of May. That makes it pointless thinking about computer upgrades (like perhaps an extra gfx card or upgrade). A pity. [Edit 1 times, last edit by DE KOMETEN at May 28, 2021 11:51:50 AM] |
||
|
|
Grumpy Swede
Master Cruncher Svíþjóð Joined: Apr 10, 2020 Post Count: 1868 Status: Offline Project Badges: 
|
Well, if they're not going to increase the available work for the 100's of times more efficient app (GPU), then I for one, is not going to participate much at all. CPU crunching when there is a GPU app available, is just a total waste of electricity.
----------------------------------------  [Edit 1 times, last edit by Grumpy Swede at May 28, 2021 12:22:23 PM] |
||
|
|
pututu
Senior Cruncher United States Joined: Jan 3, 2016 Post Count: 241 Status: Offline Project Badges:
|
I thought the task generation is like 2000 tasks instead of 1700 tasks every 30 mins as stated by uplinger here and here.
---------------------------------------- [Edit 2 times, last edit by pututu at May 29, 2021 5:37:54 AM] |
||
|
|
yoerik
Senior Cruncher Canada Joined: Mar 24, 2020 Post Count: 413 Status: Offline Project Badges:
|
I thought the task generation is like 2000 tasks instead of 1700 tasks every 30 mins as stated by uplinger here and here. it was over a month between those posts and the project update. Perhaps they found 1.7k / 30 minutes made sense given the current capabilities of the grid and the researchers' system? Hopefully they can find a way to massively increase GPUs' WU availability. |
||
|
|
Speedy51
Veteran Cruncher New Zealand Joined: Nov 4, 2005 Post Count: 1220 Status: Offline Project Badges: 
|
Everybody has their individual goals. In terms of being able to get just GPU work without babysitting at the moment to my understanding is not possible. Unless you don't want to run any CPU work
---------------------------------------- |
||
|
|
Speedy51
Veteran Cruncher New Zealand Joined: Nov 4, 2005 Post Count: 1220 Status: Offline Project Badges:
|
The project has added GPU power to the existing strong CPU power that supports research for potential COVID-19 treatments It would appear that they aren't able to process the amount of work from the CPU application let alone the GPU application as well. Results returned per day have dropped from over 1,000,00 to somewhere around the 500,000 mark. Here is the reason . From this explanation I am not sure if it is a permanent cutback or temporary, have a feeling may be permanent |
||
|
|
|