| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Active Research
Forum: OpenPandemics - COVID-19 Project
Thread: Feature request: split "Open Pandemics - COVID-19" into 2 projects - for CPU and for GPU |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 14
|
|
| Author |
|
|
sam6861
Advanced Cruncher Joined: Mar 31, 2020 Post Count: 107 Status: Offline Project Badges: 
|
I agree, split it into separate OPN1 and OPNG options in the following website places:
- My Contributions, Results Status, filters. - Device Profiles, Available projects. - Device Profiles, Project limits. If limit OpenPandemics to 1, then this will also limit OPNG to 1. Very annoying with both AMD GPU and NVidia GPU in 1 PC, where if I limit to 50, 49 on slow NVidia GT 1030, and 1 task on fast AMD RX 5500 XT, I have to use unlimited if I want lots of GPU tasks. The use of max_concurrent in app_config.xml have a bug in BOINC 7.16.11 which can cause BOINC client to get way too much work. I stopped using max_concurrent and it then work fine. As for getting the tasks, there are some current ways to get GPU only for Open Pandemic shown in some other topics. Though this may sometimes not fully work for all computers. https://www.worldcommunitygrid.org/forums/wcg/viewthread_thread,43314 |
||
|
|
Link64
Senior Cruncher Joined: Feb 19, 2021 Post Count: 206 Status: Offline Project Badges: 
|
Well, the best and also future proof posibility would be if WCG would have the "Run CPU versions of applications for which GPU versions are available" option available, this is a standard BOINC feature, they just need to enable it. No need to split OPN into two projects for such simple thing.
---------------------------------------- |
||
|
|
adriverhoef
Master Cruncher The Netherlands Joined: Apr 3, 2009 Post Count: 2346 Status: Recently Active Project Badges: 
|
In any case, there is so little OPN for GPUs (aside from the recent stress test) that you'd still get very little work. For reference, there are only 2,000 workunits release every 30 minutes. I'm very satisfied with the current supply of GPU tasks. My workunit queue has enough work for at least one day (Intel GPU). |
||
|
|
goben_2003
Advanced Cruncher Joined: Jun 16, 2006 Post Count: 146 Status: Offline Project Badges: 
|
Well, the best and also future proof posibility would be if WCG would have the "Run CPU versions of applications for which GPU versions are available" option available, this is a standard BOINC feature, they just need to enable it. No need to split OPN into two projects for such simple thing. I agree that it would be a good feature to have. Not just for now, but for future projects that hopefully will have GPU versions. I would just like to point out though that it seems like OPN is already split into two projects - everywhere except the website. WCG sends the information to BOINC as them being separate. This is why(or at least one reason) we use opn1 and opng to separately configure them and only need to use plan_class if we have multiple GPU brands. BOINC requests them separately, WCG tells the BOINC client that we have selected both <app_id>92</app_id> and <app_id>94</app_id> as well as that they have the same limits: <limited_app>  |
||
|
|
|