| 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 12 2021 [ Issues Thread ] |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 160
|
|
| Author |
|
|
goben_2003
Advanced Cruncher Joined: Jun 16, 2006 Post Count: 145 Status: Offline Project Badges: 
|
Intel® Core™ i7-8650U Processor Intel says that is a mobile, 4-core, 8-thread processor with a default TDP of 15 W - so hyperthreading is in operation for the CPU tasks. I think your results need to be considered in that context. I am well aware of the limitations of that processor.  This does effect that running tasks on both the CPU and the iGPU results in a jobs/hr slowdown of the iGPU that is more than double the total jobs/hr of just running the CPU. The CPU is throttled due to the TDP when both the CPU and iGPU are in use. It does not change the fact that: 1. The total jobs/hr is higher running 8 threads on the 4 cores than just running 4 would be. 2. The iGPU alone gets an order of magnitude more jobs/hr than the CPU can. 3. The slowdown profile is drastically different between those tests tasks and OPN. Cheers, goben_2003 Edit: Removed accidental post icon (and time for another cup of  ) [Edit 1 times, last edit by goben_2003 at Mar 14, 2021 12:41:12 PM] |
||
|
|
nanoprobe
Master Cruncher Classified Joined: Aug 29, 2008 Post Count: 2998 Status: Offline Project Badges: 
|
Edit: I have my doubts, that the project will even be able to churn out as many tasks as will be needed, when this goes into production. If my GTX980 crunces one task in 3-4 minutes (and the latest ones were tasks), then more powerful GPU's will when running multiple tasks crunch them in less than that. Multiply that with all GPU's on WCG, and there's going to be one he** of a demand. The # of CPUs here far exceeds the # of GPUs and there is plenty of work for them. I don't foresee the # of GPUs to be an issue with supply even with the faster speeds. Never happened on the HCC GPU project. There is an important difference from back then. This can run on integrated GPUs, which results in a huge increase in the number of GPUs that can run these work units. Even though they are much slower than discrete cards, they are much faster than CPUs. My RTX 2060 gets ~500 jobs/hr with only 26% load - the iGPU gets ~66 jobs/hr. That iGPU gets ~11x the jobs/hr vs all 8 threads in the i7 that it is a part of(tasks only on iGPU vs only on CPU). Whether that will result in a supply issue remains to be seen. If it does, that is a good problem to have.  On a 4c/8t cpu your iGPU would need to be 8 times faster than the cpu to do the same amount of tasks. Is it? I don’t have any iGPU mobos so I have no way to run a comparison. Even adding the iGPUs to the equation there are still considerably more cpus here than gpus.
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 14, 2021 1:36:42 PM] |
||
|
|
Paul Schlaffer
Senior Cruncher USA Joined: Jun 12, 2005 Post Count: 240 Status: Offline Project Badges: 
|
The error rate on my machines is much higher for this version. It appears that all 3 machines that have an AMD APU are erroring out, where previously only one was generating errors. The machine with the Nvidia card is validating fine.
---------------------------------------- “Where an excess of power prevails, property of no sort is duly respected. No man is safe in his opinions, his person, his faculties, or his possessions.” – James Madison (1792) |
||
|
|
Doublec
Advanced Cruncher France Joined: Aug 25, 2006 Post Count: 58 Status: Offline Project Badges:
|
Hello,
4 computers with Windows 10 Pro 64 bit (20H2). - Intel Core i7-3770K + Nvidia Geforce GTX 1070 : 22 jobs OK (pending validation) - Intel Core i7-3770 + Nvidia Geforce GTX 1650 : 8 jobs OK (7 pending validation, 1 valid) - Intel Xeon W3670 + Nvidia Quadro K620 : 8 jobs OK (pending validation) - Intel Xeon E5-1620 v3 + Nvidia Quadro K620 : 8 jobs OK (pending validation), 1 Error. Unhandled Exception Detected... - Unhandled Exception Record - Reason: Breakpoint Encountered (0x80000003) at address 0x00007FFFB7729AD2 https://www.worldcommunitygrid.org/ms/device/...og.do?resultId=1559241312 |
||
|
|
maeax
Advanced Cruncher Joined: May 2, 2007 Post Count: 142 Status: Offline Project Badges:
|
On a 4c/8t cpu your iGPU would need to be 8 times faster than the cpu to do the same amount of tasks. Is it? I don’t have any iGPU mobos so I have no way to run a comparison. Even adding the iGPUs to the equation there are still considerably more cpus here than gpus. i7-8700 with 6c/12t. Two Beta finished with 27 and 33 Calculations inside the iGPU UHD630 in 30 Min for every GPU-Task. 10 CPU-Tasks with 2 Calculations inside run every 4 hours at the same time. 10x2=20 and 27+33 = 60. So, the GPU can run in 4 hours 240 Calculations and the CPU-Tasks only 20!
AMD Ryzen Threadripper PRO 3995WX 64-Cores/ AMD Radeon (TM) Pro W6600. OS Win11pro
----------------------------------------[Edit 1 times, last edit by maeax at Mar 15, 2021 8:39:56 AM] |
||
|
|
goben_2003
Advanced Cruncher Joined: Jun 16, 2006 Post Count: 145 Status: Offline Project Badges:
|
On a 4c/8t cpu your iGPU would need to be 8 times faster than the cpu to do the same amount of tasks. Is it? I don’t have any iGPU mobos so I have no way to run a comparison. Even adding the iGPUs to the equation there are still considerably more cpus here than gpus. I am guessing you missed this in my post. The iGPU runs through jobs ~11x as fast as the combined total of all 4c/8t of the CPU. That is with just the GPU or just the CPU in use, respectively. [Edit 1 times, last edit by goben_2003 at Mar 14, 2021 8:36:34 PM] |
||
|
|
Falconet
Master Cruncher Portugal Joined: Mar 9, 2009 Post Count: 3294 Status: Offline Project Badges:
|
https://www.worldcommunitygrid.org/ms/device/...s.do?workunitId=569920905
----------------------------------------Mine is the one with 0.04 runtime. It's a Mobile AMD Vega 8 GPU from a 2500U APU. IIRC, the Vega 8 was on the list of cards erroring out a lot. I'm curious to see if it will validate. There was another Beta task but that one got the exceeded runtime limit error. 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 |
||
|
|
TLD
Veteran Cruncher USA Joined: Jul 22, 2005 Post Count: 785 Status: Offline Project Badges: 
|
Shouldn't the process feeding the GPU work unit be at a higher priority then the CPU work units.
----------------------------------------If that is the case the CPU work units would just slow down to give the process feeding the GPU work unit as much CPU power as needed.  |
||
|
|
ca05065
Senior Cruncher Joined: Dec 4, 2007 Post Count: 325 Status: Offline Project Badges: 
|
From a previous posting of mine during the Feb 18th beta test on 20/02/2021:
I have looked at the processing of the beta work units with Task Manager and Process Explorer. The PC is a Ryzen 5 3600 with a GT1030 GPU. The work units are specified as needing 0.634 of a CPU plus 1 Nvidia GPU. Process Explorer shows that the CPU part of the work unit runs with a Windows priority of 6. A CPU only work unit runs with a priority of 4 and normal windows task with a priority of 8. The CPU part used 100% of a thread while the GPU percentage jumped between low values and 100%. The work units ran in between 10 and 20 minutes with most taking about 15 minutes. |
||
|
|
Richard Haselgrove
Senior Cruncher United Kingdom Joined: Feb 19, 2021 Post Count: 360 Status: Offline Project Badges: 
|
A CPU only work unit runs with a priority of 4 and normal windows task with a priority of 8. That's interesting. Checking this morning, my CPU tasks are running with a priority of 1. That's with a generic BOINC client (v7.16.11) setting the priorities as the apps launch, under Windows 7. |
||
|
|
|