Yep. Sorry, we're still working on this. I usually deprecate new app versions as soon as the first errors are reported, took me a bit longer this time. We're still not there yet.
... used about 16% of the GPU. FGRP usually uses about 95%.
A more interesting statistic would be crunch time as compared to what a CPU can do, and of course whether they will ultimately give a close enough result for validation. It would also be interesting to know if it's possible to run two GPU tasks concurrently and if that makes any improvement.
Your GPU is a GTX 1070 and the crunch time was approximately 5500s for the two returned tasks so far.
On one of my hosts (low end Athlon 200GE) a CPU task takes ~29,700s.
For a more capable CPU, the crunch time would be somewhat lower so a rough ball-park figure is that the GPU speedup is perhaps around 4x to 5x at the moment. That's reasonable for this early stage so hopefully down the track, some optimisations might be found that increase that factor to something closer to what it is for FGRPB1G.
However, the first step really is, will they validate :-). Hopefully, they will. Your earlier failed tasks of the same frequency series have a quorum partner (an i3) but unfortunately that host doesn't seem to be returning work at the moment. Your completed tasks don't have a quorum partner but would likely get that same i3 if it suddenly started returning work and requesting more :-).
It would also be interesting to know if it's possible to run two GPU tasks concurrently and if that makes any improvement.
Yes, when I saw it only used 16%, I changed the value to the multiple task option to now 0.5, so it can try to run 2 tasks. I can increase it to 3 or more later.
I also tried running 2 FgRP taks and then releared why I choose to only run 1. With 2 tasks 99% to 100% of the GPU is used and then I cant watch videos without them being choppy.
Bernd Machenschalk
)
That version still had a bug (in handling of '--device'), so it was deprecated as soon as we noticed. New version 0.06 out now!
BM
My Windows hosts with Nvidia
)
My Windows hosts with Nvidia (GTX 960 or 1060) still can't get any "Engineering" tasks for GPU.
Hi, I posted it in bug
)
Hi, I posted it in bug reports but all the new GPU tasks I got crash immediately. Edit On this host: https://einsteinathome.org/de/host/12266191
Yep. Sorry, we're still
)
Yep. Sorry, we're still working on this. I usually deprecate new app versions as soon as the first errors are reported, took me a bit longer this time. We're still not there yet.
BM
So just a curiosity question
)
So just a curiosity question on the GW for GPUs...I see they are for windows and macs...how come no Linux?
The Version 0.11 went without
)
The Version 0.11 went without any issues, I think. And used about 16% of the GPU. FGRP usually uses about 95%.
tolafoph wrote:... used about
)
A more interesting statistic would be crunch time as compared to what a CPU can do, and of course whether they will ultimately give a close enough result for validation. It would also be interesting to know if it's possible to run two GPU tasks concurrently and if that makes any improvement.
Your GPU is a GTX 1070 and the crunch time was approximately 5500s for the two returned tasks so far.
On one of my hosts (low end Athlon 200GE) a CPU task takes ~29,700s.
For a more capable CPU, the crunch time would be somewhat lower so a rough ball-park figure is that the GPU speedup is perhaps around 4x to 5x at the moment. That's reasonable for this early stage so hopefully down the track, some optimisations might be found that increase that factor to something closer to what it is for FGRPB1G.
However, the first step really is, will they validate :-). Hopefully, they will. Your earlier failed tasks of the same frequency series have a quorum partner (an i3) but unfortunately that host doesn't seem to be returning work at the moment. Your completed tasks don't have a quorum partner but would likely get that same i3 if it suddenly started returning work and requesting more :-).
Cheers,
Gary.
Gary Roberts schrieb: It
)
Yes, when I saw it only used 16%, I changed the value to the multiple task option to now 0.5, so it can try to run 2 tasks. I can increase it to 3 or more later.
I also tried running 2 FgRP taks and then releared why I choose to only run 1. With 2 tasks 99% to 100% of the GPU is used and then I cant watch videos without them being choppy.
Got a
)
Got a validation!
https://einsteinathome.org/workunit/396688820
That's very good to know!
)
That's very good to know! Your 1050Ti outperformed (one core of) a decent i7 by a factor of 4.53x :-).
I wonder how much better a 1080Ti might be :-). Not to mention how a decent AMD GPU might go :-).
Cheers,
Gary.