The OS is OpenSUSE 15.4, using the amdgpu driver. clinfo shows the following:
Device Name gfx1010:xnack-
Device Vendor Advanced Micro Devices, Inc.
Device Vendor ID 0x1002
Device Version OpenCL 2.0
Driver Version 3452.0 (HSA1.1,LC)
Device OpenCL C Version OpenCL C 2.0
Device Type GPU
Device Board Name (AMD) AMD Radeon RX 5700 XT
The time yours gpu tasks are taking vary widely.
What percent of threads/CPU and percent of time for each CPU are you running.
An overcommitted CPU could be an issue.
I usually have 2-4 CPUs/threads not committed to boinc. While running 100 percent of my CPU time.
I also have boinc suspend if non-boinc tasks exceed some percent.
Hth,
Tom M
A Proud member of the O.F.A. (Old Farts Association). Be well, do good work, and keep in touch.® (Garrison Keillor) I want some more patience. RIGHT NOW!
I'm running BOINC in a virtual machine. The client is allowed 100% of the cores and CPU time, since the virtual machine already segments the resources I'm willing to commit to BOINC. The host is at about 50% CPU and RAM usage.
I never had any problems running the FGRPopencl1K-ati tasks using the same machine and settings.
I'm running BOINC in a virtual machine. The client is allowed 100% of the cores and CPU time, since the virtual machine already segments the resources I'm willing to commit to BOINC. The host is at about 50% CPU and RAM usage.
I never had any problems running the FGRPopencl1K-ati tasks using the same machine and settings.
I am sorry I have no further ideas. Probably going back to the FGRPopencl1K-ati is the best idea.
Tom M
A Proud member of the O.F.A. (Old Farts Association). Be well, do good work, and keep in touch.® (Garrison Keillor) I want some more patience. RIGHT NOW!
I'm running BOINC in a virtual machine. The client is allowed 100% of the cores and CPU time, since the virtual machine already segments the resources I'm willing to commit to BOINC. The host is at about 50% CPU and RAM usage.
I never had any problems running the FGRPopencl1K-ati tasks using the same machine and settings.
What other Boinc cpu Projects are you running and how many Einstein cpu tasks are you running, it sounds like the cpu is busy doing something else and when you force it to recognize Einstein again it says 'oh yeah I remember what I'm supposed to be doing'. Remember too Einstein always reserves a full cpu core just for the gpu to use are you by chance bypassing that? How many gpu tasks are you running at the same time?
As far as not having trouble running the GRP tasks remember each type of task is different and has different variables and parameters when they make them. So that's a bit like me saying I can run Universe tasks why can't I run LHC tasks...because they are VERY VERY different not counting that they are from different Projects.
firebringer11 wrote: My
)
The time yours gpu tasks are taking vary widely.
What percent of threads/CPU and percent of time for each CPU are you running.
An overcommitted CPU could be an issue.
I usually have 2-4 CPUs/threads not committed to boinc. While running 100 percent of my CPU time.
I also have boinc suspend if non-boinc tasks exceed some percent.
Hth,
Tom M
A Proud member of the O.F.A. (Old Farts Association). Be well, do good work, and keep in touch.® (Garrison Keillor) I want some more patience. RIGHT NOW!
I'm running BOINC in a
)
I'm running BOINC in a virtual machine. The client is allowed 100% of the cores and CPU time, since the virtual machine already segments the resources I'm willing to commit to BOINC. The host is at about 50% CPU and RAM usage.
I never had any problems running the FGRPopencl1K-ati tasks using the same machine and settings.
firebringer11 wrote: I'm
)
I am sorry I have no further ideas. Probably going back to the FGRPopencl1K-ati is the best idea.
Tom M
A Proud member of the O.F.A. (Old Farts Association). Be well, do good work, and keep in touch.® (Garrison Keillor) I want some more patience. RIGHT NOW!
firebringer11 wrote:I'm
)
What other Boinc cpu Projects are you running and how many Einstein cpu tasks are you running, it sounds like the cpu is busy doing something else and when you force it to recognize Einstein again it says 'oh yeah I remember what I'm supposed to be doing'. Remember too Einstein always reserves a full cpu core just for the gpu to use are you by chance bypassing that? How many gpu tasks are you running at the same time?
As far as not having trouble running the GRP tasks remember each type of task is different and has different variables and parameters when they make them. So that's a bit like me saying I can run Universe tasks why can't I run LHC tasks...because they are VERY VERY different not counting that they are from different Projects.
mikey wrote: What other
)
Look into his BOINCstats listing and you'll see that he's running 5 of 9 Projects.
Proud member of the Old Farts Association
GWGeorge007 wrote: mikey
)
That could be the problem then
FGRPB1G: Tasks to send –
)
FGRPB1G: Tasks to send – 0
Are we running out of FGRPB1G tasks now?
Yes. Switch the BRP7
)
Yes. Switch to BRP7
_________________________________________________________________________
There seems to be a lot of
)
There seems to be a lot of FGRP5 (CPUs?) to do. Can this not be moved over to graphics cards?
If this page takes an hour to load, reduce posts per page to 20 in your settings, then the tinpot 486 Einstein uses can handle it.
TRAPPIST-713 wrote:Are we
)
The most recent information about this is here.
Cheers,
Gary.