It works on the O3 tasks, at least I got thru it okay!! One task at a time, more tasks to come tomorrow after I finish up another Project
h1_0474.20_O3aC01Cl1In0__O3MDFG1_G34731_474.50Hz_570_1 | 700766350 | 10 Jan 2023 3:17:17 UTC | 10 Jan 2023 3:27:17 UTC | Completed, waiting for validation | 557 | 363 | 0 |
Multi-Directional Gravitational Wave search on O3
|
https://einsteinathome.org/host/12541707/tasks/0/0
mikey
Copyright © 2024 Einstein@Home. All rights reserved.
Sounds encouraging!
)
Sounds encouraging!
A Proud member of the O.F.A. (Old Farts Association). Be well, do good work, and keep in touch.® (Garrison Keillor)
Tom M wrote: Sounds
)
It is thanks, I use MSIAfterburner and it's saying the gpu usage is in high 70's to the low 80's so I think I could run more than 1 task at a time but I want it to run for a week or so before I think about doing that. Einstein needs move venues as I've run out!!
mikey wrote: Tom M
)
What are each venue setup to do? Would it help on GPU crunching to setup a specific task and then toggle the "when task not available, run available"?
Just curious. I have had to reuse profiles depending on what I was trying to do.
Tom M
A Proud member of the O.F.A. (Old Farts Association). Be well, do good work, and keep in touch.® (Garrison Keillor)
Tom M wrote: mikey
)
I have generic setup as zero resource share cpu only, I use home as a 2nd cpu only venue but with a non zero number for the resource share. I use school as a zero resource share gpu only venue and work as a non zero resource share gpu project. That way if I'm just coming over to a project with a pc that hasn't run it in awhile, or running new types of tasks I use the zero resource share venue so I don't get swamped with ie 300 tasks that then error out because of some problem of other. Once I figure out that the tasks can run okay I will move them to the non zero resource share venue for long term crunching so they keep getting new tasks in batches, or one at a time but they keep the cache filled. Whereas the zero resource share venue tends to only get enough tasks to run right now with a couple of spares at the ready not a whole day or so worth of tasks.
Tom M wrote: Sounds
)
The tasks are now all ending in errors so i am stopping them and I'm going to try a less intense gpu project like Milkyway. I've rebooted a couple of times and it kinda sorta works but if I go to bed at night and everything I get is errors then it can't run it here, I have too many machines to babysit one of them like that.
mikey wrote: Tom M
)
It seems I can do the Meerkat and BRP7 tasks on it but not the GRP#1 tasks for some reason, I wonder if the cards memory is giving me troubles for those tasks?
it might be that that card
)
it might be that that card has something wrong with it.
or a problem with the drivers.
_________________________________________________________________________
Ian&Steve C. wrote:it might
)
I agree, I don't know how it was used before I got it but it was REALLY clean, that made me think of mining but who really knows. The person shipped it the very next day after I bought it, just beyond my local pickup range, and I got it the next day.
The drivers are the ones suggested by AMD when I had them scan what I have in the pc,
AMD Adrenalin ver 22.11.2
What does BOINC report for
)
What does BOINC report for the card capabilities in the beginning of the Event Log?
The details on the card on the website don't report anything other than the card name.
I assume that the ATI version of the 1.28 application also needs OpenCL 2.0 capabilities to run correctly.
What does clinfo or the Windows equivalent report for its OpenCL capability?
Oh I forgot about that and
)
Oh I forgot about that and didn’t scrutinize.
the v1.28 app has never worked for Vega on windows and the AMD v1.28 app has been in beta. Something wrong with the windows drivers not providing full support.
mikey, if you disable beta tasks then you’ll get the older app that should work.
_________________________________________________________________________