FYI the problem happening in SETI to, so is not E@H related it´s something diferent.
For all i was ablle to dig it´s related to the driver itself, but the older drivers seems to be slower so the only real option for now it´s running 1 wu at a time.
What i did not try is what happening when you start 2 compleately separate boincs on the same host (not just run 2 wu at a time), may that works but i´m unable to test that option remotely maybe you could try and tell us what happeing.
Are saying to run two totally separate instances of Boinc on the SAME gpu?
Yes, there is a way to run simultaneus boinc instances on the same host. I do that in the past to bypass the 100 WU limit on SETI and works perfect. Since each instances works as a single WU cruncher (even if they use the same GPU) i belive that could could be a way to fix the problem.
My ideia is to try that and see if the problem remains or not.
yep, it should work, because when i gaming on my pc with e@h running in background i still receive no invalid results. so games does not affect e@h, so second instance of boinc should work.
but i can't try it soon because i'm away from pc for a couple of weeks )
and i don't know how to run second instance of boinc (
That´s exactly why i believe that could work, unfortunately i can´t try because i only access my cruncher (i´m in a diferent country) remotely and configurate to do that remotely is not an easy task.
But if somebody wish to try i could try to guide how to do.
I'm pretty sure my R9 390 was crunching 2 WU at a time. If I was only running E@H tasks I could leave my CPU on 100% and set the BRP preferences to 0.5.
Having said that.... Im not seeing ANY GPU WU's now :/
Hope this isn't the same issue as MW@H where I can't do any GPU crunching without writing all this code :/
I'm pretty sure my R9 390 was crunching 2 WU at a time.
It looks very much as if you were judging by the huge amount of tasks you've completed that have resulted in validate errors ;-)
Once the flow of work is restored I'd advise running BRP's singly for maximum (reliable) throughput... There is an as yet undiscovered issue with 'Hawaii' onward AMD GPU's wrt running multiple tasks and from what I've read the problem is affecting all of the better known projects.
I don't subscribe to the thought that its a driver problem per se, many people are happily using the latest drivers without issue with older GPU varieties.
I suspect (with Hawaii onwards) there is an incompatibility somewhere in the code that doesn't like OpenCL 2 devices or vice versa (regardless of manufacturer claimed backwards compatibility).
What we need is to prick the interest of the project dev's into looking at this from the projects app code point of view to narrow down the list of potential culprits causing tasks to fail when run concurrently.
Sadly, I doubt there are enough Hawaii and to a certain degree Tonga users here at Einstein (wanting to run multiple tasks) to make it worth the developers effort.
RE: FYI the problem
)
Are saying to run two totally separate instances of Boinc on the SAME gpu?
Yes, there is a way to run
)
Yes, there is a way to run simultaneus boinc instances on the same host. I do that in the past to bypass the 100 WU limit on SETI and works perfect. Since each instances works as a single WU cruncher (even if they use the same GPU) i belive that could could be a way to fix the problem.
My ideia is to try that and see if the problem remains or not.
yep, it should work, because
)
yep, it should work, because when i gaming on my pc with e@h running in background i still receive no invalid results. so games does not affect e@h, so second instance of boinc should work.
but i can't try it soon because i'm away from pc for a couple of weeks )
and i don't know how to run second instance of boinc (
That´s exactly why i believe
)
That´s exactly why i believe that could work, unfortunately i can´t try because i only access my cruncher (i´m in a diferent country) remotely and configurate to do that remotely is not an easy task.
But if somebody wish to try i could try to guide how to do.
i wish
)
i wish
You could locate the
)
You could locate the instrunctions how to run 2 Boinc at the same time in:
http://vyper.kafit.se/wp/index.php/2011/02/04/running-different-nvidia-architectures-most-optimal-at-setihome/
I'm pretty sure my R9 390 was
)
I'm pretty sure my R9 390 was crunching 2 WU at a time. If I was only running E@H tasks I could leave my CPU on 100% and set the BRP preferences to 0.5.
Having said that.... Im not seeing ANY GPU WU's now :/
Hope this isn't the same issue as MW@H where I can't do any GPU crunching without writing all this code :/
Edit - I see they have run out of WU's. Bummer
RE: I'm pretty sure my R9
)
It looks very much as if you were judging by the huge amount of tasks you've completed that have resulted in validate errors ;-)
Once the flow of work is restored I'd advise running BRP's singly for maximum (reliable) throughput... There is an as yet undiscovered issue with 'Hawaii' onward AMD GPU's wrt running multiple tasks and from what I've read the problem is affecting all of the better known projects.
I don't subscribe to the thought that its a driver problem per se, many people are happily using the latest drivers without issue with older GPU varieties.
I suspect (with Hawaii onwards) there is an incompatibility somewhere in the code that doesn't like OpenCL 2 devices or vice versa (regardless of manufacturer claimed backwards compatibility).
What we need is to prick the interest of the project dev's into looking at this from the projects app code point of view to narrow down the list of potential culprits causing tasks to fail when run concurrently.
Sadly, I doubt there are enough Hawaii and to a certain degree Tonga users here at Einstein (wanting to run multiple tasks) to make it worth the developers effort.
Perhaps we should start a petition!!
Yeah Im really bummed at the
)
Yeah Im really bummed at the moment.
No Milky way WU's, No Einstein WU's I'm stuck doing Seti@Home for GPU work when Id rather do the other 2.
Got this nice, new shiny card and now no work for it :(
I have 4 AMD Radeon R9 Fury X
)
I have 4 AMD Radeon R9 Fury X VGA cards in one PC.
Windows 8.1 Pro x64.
AMD Catalyst v15.9.1 Beta.
Just 1 WU/GPU possible - errors with 2+WUs/GPU.
Someone tried already this 'two BOINC Clients thing'?
I worry this wouldn't work, because then there are also 2 WUs/GPU...
Thanks.