Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
World Community Grid Forums
Category: Beta Testing Forum: Beta Test Support Forum Thread: OpenPandemics GPU Beta Test - March 26 2021 [ Issues Thread ] |
No member browsing this thread |
Thread Status: Locked Total posts in this thread: 511
|
Author |
|
nanoprobe
Master Cruncher Classified Joined: Aug 29, 2008 Post Count: 2998 Status: Offline Project Badges: |
There are a few things I may want to try....One of which would be a target for fixing the GT cards on NVIDIA...The other is, I want to atleast put in place a point system for the next round of beta...ya'll deserve credit and time that your systems are computing work units. Keith, just like the Nvidia issues we had with the HCC app there is probably nothing on your end that can be done for the old GT cards. Many of them just don't support OpenCl 1.2 despite what the web or output files say. Hope you can find something but I also hope you don't waste too much time on it. I don't know how much Nvidia has upped their game for OpenCl support but I do know it was dreadful when the HCC app was running. The newer cards seem to work well but I still think they are far more interested in their CUDA app than OpenCl. JMHO Good luck. I'm sure you'll be asked that if you come up with a points allocation system for the new betas will it be retroactive? Thought I'd just get that out of the way. 1k points each is fair.
In 1969 I took an oath to defend and protect the U S Constitution against all enemies, both foreign and Domestic. There was no expiration date.
----------------------------------------[Edit 1 times, last edit by nanoprobe at Mar 28, 2021 7:27:44 PM] |
||
|
Richard Haselgrove
Senior Cruncher United Kingdom Joined: Feb 19, 2021 Post Count: 360 Status: Offline Project Badges: |
I don't know what core client your using that writes it's own xml files but I would not use it under any circumstances. I know of no BOINC code that was written specifically for BT so I'll have to take you word for that. More than likely the horribly written code is your problem so why use it? I don't understand why someone is trying to reinvent the wheel. Stand alone BOINC with the BT app has worked fine for a long time. My suggestion would be to install the most up to date version of BOINC from the Berkley site, write your own xml files as needed and dump whatever it is you're trying to make work here. Good luck. I was using the stock Windows v7.16.11 - the nearest we have to a current reference release. The linkage was added to the BOINC master code-base in May 2017, by David Anderson - Fred (Efmer) added it to BoincTasks on the assumption that David's code is usable. It isn't. All the references are sourced in various posts in this sub-topic, scattered through the thread.I'm going to try and write a patch tomorrow, so that in future releases the 'edit app_config.xml' feature works as David and Fred intended. But in the meantime, it's broken - don't use it. Write you own app_config.xml file from the user documentation instead. [I was asked to act as Release Manager for v7.10.2 of BOINC - the first to include this function. I missed the bug and let it through, so I feel I have some responsibility for this fiasco. Sore point.] [Edit 1 times, last edit by Richard Haselgrove at Mar 28, 2021 7:36:24 PM] |
||
|
nanoprobe
Master Cruncher Classified Joined: Aug 29, 2008 Post Count: 2998 Status: Offline Project Badges: |
@ Richard
----------------------------------------Now I have a better picture of what's going on. I've never even attempted to use the edit config files portion of BT so I was a little confused as to what you were trying to attempt. Hope you can sort it out for your own satisfaction but I'm just too old school and prefer the old method. Have fun.
In 1969 I took an oath to defend and protect the U S Constitution against all enemies, both foreign and Domestic. There was no expiration date.
|
||
|
robertmiles
Senior Cruncher US Joined: Apr 16, 2008 Post Count: 443 Status: Offline Project Badges: |
I've found that some of the older Nvidia-based cards did not always have a specific GPU. For example, the GT 545 was sometimes built with a Fermi-generation GPU, and sometimes with the generation after Fermi. This means that not all of the GT 545 cards had the same specifications of what they could do. I have not found similar information on whether this is also true for the more recent cards.
|
||
|
Richard Haselgrove
Senior Cruncher United Kingdom Joined: Feb 19, 2021 Post Count: 360 Status: Offline Project Badges: |
I have not found similar information on whether this is also true for the more recent cards. Wikipedia is not authoritative, but has a comprehensive list.https://en.wikipedia.org/wiki/List_of_Nvidia_graphics_processing_units |
||
|
widdershins
Veteran Cruncher Scotland Joined: Apr 30, 2007 Post Count: 674 Status: Offline Project Badges: |
I never did get the GT730 to complete any WU without errors. Tried various drivers, no difference. It looks like something in the code is trying to access an area of memory it shouldn't, I have no idea what the differences between CL1.1 and 1.2 are as I'm not a coder. Is it possible that the card is generally CL1.2 feature compatible, but something in the way memory can be accessed by CL1.2 is incompatible with certain NVidia memory configurations?
On the subject of allocating points, I'd suggest the simplest way is to calculate the average point per ligand for cpu only tasks and then give the same credit for each GPU task. Won't be perfect, since the credit is not specific to the difficulty of that particular WU, but better than zero. Since it is based on work completed rather than time elapsed there would be no gaming the system with slower GPU's. And finally though the credit would not reflect that individual unit's difficulty, since GPU's would be processing 100's of units per day it'd all even out when he same credit per ligand were given to easier WU's. |
||
|
poppinfresh99
Cruncher Joined: Feb 29, 2020 Post Count: 49 Status: Offline Project Badges: |
I have an Intel-GPU laptop (dell latitude 3340 from 2014 I think) that is seeing an error after 12.42 hours on each task (0.01 hours of CPU time). My other Intel device was working just fine. Since it received at least 9 tasks, I will have to abort many of them.
Maybe somebody mentioned this Intel issue already, but there are 19 pages in this forum thread, which is about 17 too many for me to read. <core_client_version>7.16.11</core_client_version> |
||
|
PMH_UK
Veteran Cruncher UK Joined: Apr 26, 2007 Post Count: 761 Status: Offline Project Badges: |
All units failed (hung) on my oldish Windows 7 PC with Intel GPU but similar (identical?) PC running Linux is OK, valid units.
----------------------------------------Updated GPU driver to Intel latest but no change. Paul.
Paul.
|
||
|
ca05065
Senior Cruncher Joined: Dec 4, 2007 Post Count: 325 Status: Offline Project Badges: |
During the first two rounds of the GPU beta testing, I received 30 or 40 work units on my Ryzen 5 3600 with an Nvidia GT1030 video card. They were all valid. Since then I have not received any GPU work units.
During BOINC start it can see the video card: 26-Mar-2021 05:58:34 [---] Running under account colinc 26-Mar-2021 05:58:34 [---] CUDA: NVIDIA GPU 0: GeForce GT 1030 (driver version 461.09, CUDA version 11.2, compute capability 6.1, 2048MB, 1661MB available, 1127 GFLOPS peak) 26-Mar-2021 05:58:34 [---] OpenCL: NVIDIA GPU 0: GeForce GT 1030 (driver version 461.09, device version OpenCL 1.2 CUDA, 2048MB, 1661MB available, 1127 GFLOPS peak) 26-Mar-2021 05:58:34 [---] Windows processor group 0: 12 processors 26-Mar-2021 05:58:34 [---] Host name: DESKTOP-0AUP2S8 26-Mar-2021 05:58:34 [---] Processor: 12 AuthenticAMD AMD Ryzen 5 3600 6-Core Processor [Family 23 Model 113 Stepping 0] When requesting work units I see: 26-Mar-2021 06:00:37 [World Community Grid] [sched_op] Starting scheduler request 26-Mar-2021 06:00:38 [World Community Grid] Sending scheduler request: To fetch work. 26-Mar-2021 06:00:38 [World Community Grid] Requesting new tasks for CPU 26-Mar-2021 06:00:38 [World Community Grid] [sched_op] CPU work request: 3632202.95 seconds; 0.00 devices 26-Mar-2021 06:00:38 [World Community Grid] [sched_op] NVIDIA GPU work request: 0.00 seconds; 0.00 devices 26-Mar-2021 06:00:40 [World Community Grid] Scheduler request completed: got 0 new tasks 26-Mar-2021 06:00:40 [World Community Grid] [sched_op] Server version 701 26-Mar-2021 06:00:40 [World Community Grid] No tasks sent 26-Mar-2021 06:00:40 [World Community Grid] No tasks are available for OpenPandemics - COVID 19 26-Mar-2021 06:00:40 [World Community Grid] No tasks are available for the applications you have selected. 26-Mar-2021 06:00:40 [World Community Grid] Tasks for AMD/ATI GPU are available, but your preferences are set to not accept them 26-Mar-2021 06:00:40 [World Community Grid] Tasks for Intel GPU are available, but your preferences are set to not accept them 26-Mar-2021 06:00:40 [World Community Grid] Project requested delay of 121 seconds 26-Mar-2021 06:00:40 [World Community Grid] [sched_op] Deferring communication for 00:02:01 26-Mar-2021 06:00:40 [World Community Grid] [sched_op] Reason: requested by project In the website settings: Beta participation is selected GPU settings: Do work on my graphics card while computer is in use? YES Use my AMD/ATI graphics card if possible: YES Use my Intel graphics card if possible: YES Use my NVIDIA graphics card if possible YES When I was receiving GPU work units I saw: 26-Feb-2021 06:55:37 [World Community Grid] [sched_op] Starting scheduler request 26-Feb-2021 06:55:38 [World Community Grid] Sending scheduler request: To fetch work. 26-Feb-2021 06:55:38 [World Community Grid] Requesting new tasks for CPU and NVIDIA GPU 26-Feb-2021 06:55:38 [World Community Grid] [sched_op] CPU work request: 3658302.46 seconds; 0.00 devices 26-Feb-2021 06:55:38 [World Community Grid] [sched_op] NVIDIA GPU work request: 345600.00 seconds; 1.00 devices 26-Feb-2021 06:55:40 [World Community Grid] Scheduler request completed: got 0 new tasks 26-Feb-2021 06:55:40 [World Community Grid] [sched_op] Server version 701 26-Feb-2021 06:55:40 [World Community Grid] No tasks sent 26-Feb-2021 06:55:40 [World Community Grid] No tasks are available for OpenPandemics - COVID 19 26-Feb-2021 06:55:40 [World Community Grid] No tasks are available for Mapping Cancer Markers 26-Feb-2021 06:55:40 [World Community Grid] No tasks are available for the applications you have selected. 26-Feb-2021 06:55:40 [World Community Grid] Tasks for AMD/ATI GPU are available, but your preferences are set to not accept them 26-Feb-2021 06:55:40 [World Community Grid] Tasks for Intel GPU are available, but your preferences are set to not accept them 26-Feb-2021 06:55:40 [World Community Grid] Project requested delay of 121 seconds 26-Feb-2021 06:55:41 [World Community Grid] [sched_op] Deferring communication for 00:02:01 26-Feb-2021 06:55:41 [World Community Grid] [sched_op] Reason: requested by project The difference is that it says it is requesting work for CPU and GPU and says how much of each it wants (I limit the number of work units downloaded using the ‘project limits’ section on the website). I am looking for suggestions of what else I need to check as I do not remember changing anything deliberately. |
||
|
PMH_UK
Veteran Cruncher UK Joined: Apr 26, 2007 Post Count: 761 Status: Offline Project Badges: |
Check local preferences are not set.
----------------------------------------If using web setting then Options, Computing Preferences shows: "Using web-based preferences from <web-site>" Paul.
Paul.
|
||
|
|