Any suggestions how to ensure or enable or tweak my Intel i5 and HD 4000 drivers are *happening here? All I've done so far is enable it in BIOS and Device Mgr but don't know what else to do about it - it doesn't seem to show in any computer config in BOINC. Would just plugging a DVI cable in it but not my monitor avail anything?(referencing a comment I saw about cheating it with a Dongle) My little 8400 GS beats the pants off it on my Desktop and what I do. Just want the HD 4000 available if anything BOINC wants it.
The way I've been recommending is to just extend the desktop onto the 2nd GPU, force use VGA, then restart Boinc,
you can connect a monitor to the HD 4000 but I don't know how reliable that is.
OK well, NVIDIA OpenGL detected problem with Driver and cannot continue message twice after 2 hrs each instance; and odd behavior of task config/runs re: kinda over-amping my Resources prefs. So I just Suspended it and zero'd the resources so SETI could get back to it unhampered. Hardware updates Monday afternoon after card gets here, then I'll try again after I get it settled and confirm freshest Drivers etc.
You can just use the Activity menu to set GPU to Suspend GPU, then Einstein CPU tasks still can progress.
Thanks Claggy. I suspended GPU @Einstein and Resumed. Resumed Rosetta too (although all their queued jobs are 5 hr versions)-so get some crunching done over there till card gets here tomorrow. Thinking I'll skip worrying about Intel GPU but thanks for the suggestions. I'll keep it enabled but not cabled. Windows and Boinc can sort it out.
Thanks Claggy. I suspended GPU @Einstein and Resumed. Resumed Rosetta too (although all their queued jobs are 5 hr versions)-so get some crunching done over there till card gets here tomorrow. Thinking I'll skip worrying about Intel GPU but thanks for the suggestions. I'll keep it enabled but not cabled. Windows and Boinc can sort it out.
Claggy can tell you for sure but if it is a different model gpu you may find all your existing gpu units aborted anyway. Some projects send units to a specific gpu model, any change and you will just get new units to replace your current one, I do NOT remember if Einstein does that or not as I haven't changed a gpu in a while.
.I do NOT remember if Einstein does that or not as I haven't changed a gpu in a while.
I was changing a GPU from GT 210 to GTX 770, 560 Ti, Quadro CX to get some benchmarks. The work unit was successfully validated being processed by different GPUs.
.I do NOT remember if Einstein does that or not as I haven't changed a gpu in a while.
I was changing a GPU from GT 210 to GTX 770, 560 Ti, Quadro CX to get some benchmarks. The work unit was successfully validated being processed by different GPUs.
I recently shuffled my GPU population in getting to my current configuration and conducting some performance comparisons. I am pretty sure that during that exercise I had more than one Perseus GPU WU which started on one type and ended on another, successfully. The GPU types in question included GTX 660, GTX 750, GTX 750 Ti, and GTX 970.
I did not observe any early terminations or failures which seemed to result from the type discrepancy.
All the hosts were Windows 7, and the only type of GPU work enabled was Einstein Perseus Arm Survey. I have no information on other cases.
Thanks Sid and mikey and Claggy for the input/feedback. Looking into Intel OpenCL SDK as a means to update my i5 CPU. I *think it's already OpenCL 1.1 and 1.2 and SSE "plenty" to coop with Einstein and SETI apps. But considering running the thing anyway to be sure. I'm fuzzy about OpenGL specs/capabilities and noticed NVIDIA whining about it but my new GT 640 will be here shortly and hopefully that'll resolve these issues via(no pun) updated NV drivers. Dunno if CPU and HD GFX driver updates will be of any consequence or benefit, just doing what I *can. In the meantime, been happily crunchin away @ SETI.
Thanks Claggy. I suspended GPU @Einstein and Resumed. Resumed Rosetta too (although all their queued jobs are 5 hr versions)-so get some crunching done over there till card gets here tomorrow. Thinking I'll skip worrying about Intel GPU but thanks for the suggestions. I'll keep it enabled but not cabled. Windows and Boinc can sort it out.
Claggy can tell you for sure but if it is a different model gpu you may find all your existing gpu units aborted anyway. Some projects send units to a specific gpu model, any change and you will just get new units to replace your current one, I do NOT remember if Einstein does that or not as I haven't changed a gpu in a while.
Einstein only has one GPU app (for each vendor) per work type, it isn't as if the project has multiple apps of different Cuda versions, where earlier Cuda version apps may not produce the correct app result on post-fermi GPUs.
Looking into Intel OpenCL SDK as a means to update my i5 CPU. I *think it's already OpenCL 1.1 and 1.2 and SSE "plenty" to coop with Einstein and SETI apps. But considering running the thing anyway to be sure. I'm fuzzy about OpenGL specs/capabilities and noticed NVIDIA whining about it but my new GT 640 will be here shortly and hopefully that'll resolve these issues via(no pun) updated NV drivers. Dunno if CPU and HD GFX driver updates will be of any consequence or benefit, just doing what I *can. In the meantime, been happily crunchin away @ SETI.
For correct Intel GPU operation on this project, you need an Intel GPU OpenCL 1.2 capable driver, no later than 15.33.22.64.3621, ie not the latest, they produce inconclusive/invalid results,
there are download urls in the BRP4 Intel GPU app feedback thread thread for that last working driver, earlier OpenCL 1.1 drivers are no good eithier.
RE: Any suggestions how to
)
The way I've been recommending is to just extend the desktop onto the 2nd GPU, force use VGA, then restart Boinc,
you can connect a monitor to the HD 4000 but I don't know how reliable that is.
Claggy
RE: OK well, NVIDIA OpenGL
)
You can just use the Activity menu to set GPU to Suspend GPU, then Einstein CPU tasks still can progress.
Claggy
Thanks Claggy. I suspended
)
Thanks Claggy. I suspended GPU @Einstein and Resumed. Resumed Rosetta too (although all their queued jobs are 5 hr versions)-so get some crunching done over there till card gets here tomorrow. Thinking I'll skip worrying about Intel GPU but thanks for the suggestions. I'll keep it enabled but not cabled. Windows and Boinc can sort it out.
RE: Thanks Claggy. I
)
Claggy can tell you for sure but if it is a different model gpu you may find all your existing gpu units aborted anyway. Some projects send units to a specific gpu model, any change and you will just get new units to replace your current one, I do NOT remember if Einstein does that or not as I haven't changed a gpu in a while.
RE: .I do NOT remember if
)
I was changing a GPU from GT 210 to GTX 770, 560 Ti, Quadro CX to get some benchmarks. The work unit was successfully validated being processed by different GPUs.
RE: .I do NOT remember if
)
I was changing a GPU from GT 210 to GTX 770, 560 Ti, Quadro CX to get some benchmarks. The work unit was successfully validated being processed by different GPUs.
I recently shuffled my GPU
)
I recently shuffled my GPU population in getting to my current configuration and conducting some performance comparisons. I am pretty sure that during that exercise I had more than one Perseus GPU WU which started on one type and ended on another, successfully. The GPU types in question included GTX 660, GTX 750, GTX 750 Ti, and GTX 970.
I did not observe any early terminations or failures which seemed to result from the type discrepancy.
All the hosts were Windows 7, and the only type of GPU work enabled was Einstein Perseus Arm Survey. I have no information on other cases.
Thanks Sid and mikey and
)
Thanks Sid and mikey and Claggy for the input/feedback. Looking into Intel OpenCL SDK as a means to update my i5 CPU. I *think it's already OpenCL 1.1 and 1.2 and SSE "plenty" to coop with Einstein and SETI apps. But considering running the thing anyway to be sure. I'm fuzzy about OpenGL specs/capabilities and noticed NVIDIA whining about it but my new GT 640 will be here shortly and hopefully that'll resolve these issues via(no pun) updated NV drivers. Dunno if CPU and HD GFX driver updates will be of any consequence or benefit, just doing what I *can. In the meantime, been happily crunchin away @ SETI.
RE: RE: Thanks Claggy. I
)
Einstein only has one GPU app (for each vendor) per work type, it isn't as if the project has multiple apps of different Cuda versions, where earlier Cuda version apps may not produce the correct app result on post-fermi GPUs.
Claggy
RE: Looking into Intel
)
For correct Intel GPU operation on this project, you need an Intel GPU OpenCL 1.2 capable driver, no later than 15.33.22.64.3621, ie not the latest, they produce inconclusive/invalid results,
there are download urls in the BRP4 Intel GPU app feedback thread thread for that last working driver, earlier OpenCL 1.1 drivers are no good eithier.
Claggy