GPU72 and BOINC a match made in ....

[rebirther]

Originally Posted by KEP

The ini file, for mfact(o)(c) needs to be:

StopAfterFactor=2

Its not much of a deal, but it will speed up processing a little more, since we achieved the goal of finding a factor and eliminating a composite test.

Nice work, with a whopping 103 hosts currently connected

mfakto has it, I will change it for mfaktc too, Linux cuda seems to be working, need to make a list which are not working

[KEP]

Originally Posted by rebirther

mfakto has it, I will change it for mfaktc too, Linux cuda seems to be working, need to make a list which are not working

That makes sense, since my GPU is an AMD, good job Reb

So Linux is catching up and at least windows apps is working without problems, that is afterall where most users is working from

[rebirther]

The most common error on linux opencl with AMD is ERROR: unknown option '--device' GPU not supported?

[chalsall]

So, the results from the first batch are starting to come in!

Exactly the expected number of factors found. Good job sir!

[LaurV]

Originally Posted by chalsall

Good job sir!

Yarrrr!
Already in TF Top500 Life Time, with about 20k GHzDays in 2 days (position ~430, last reported results not yet spidered by gpu72).

With this output we're starting fearing for our skin, and we think we need to add some more of our cards to the fry... (and we assume that the boinc users didn't start yet to pour in!)

Thumb up guys!

[pinhodecarlos]

Originally Posted by LaurV

Yarrrr!
Already in TF Top500 Life Time, with about 20k GHzDays in 2 days (position ~430, last reported results not yet spidered by gpu72).

With this output we're starting fearing for our skin, and we think we need to add some more of our cards to the fry... (and we assume that the boinc users didn't start yet to pour in!)

Thumb up guys!

You won’t have a chance. When the client gets stable you will see at least an increase of the output by 15x-20x or even more. The clients connected right now are a small percentage and the usual guys when a new BOINC app pops out on this world, first guys to find about new apps like guys from Team Anandtech, Seti.USA, UK overclokers and etc. I would say in a month time you will the outcome.

[KEP]

Originally Posted by pinhodecarlos

You won’t have a chance. When the client gets stable you will see at least an increase of the output by 15x-20x or even more. The clients connected right now are a small percentage and the usual guys when a new BOINC app pops out on this world, first guys to find about new apps like guys from Team Anandtech, Seti.USA, UK overclokers and etc. I would say in a month time you will the outcome.

Maybe some of the users active at SETI@home, should refer the guys there and their attention to putting their GPU to good use, now that SETI@home stops feeding their GPUs after march 31st. If all the SETI@home users moved their GPUs to srbase to help GIMPs, that would really blow the ceiling and send us beyond the Moon - compared to the firepower we already see.

Have you begun spreading the word about GPU72?

[LaurV]

Originally Posted by pinhodecarlos

You won’t have a chance.

I know. That was a joke. In spite of the fact that my actual output (about 8THzD/D) is about 1/6 of the total computing power I could pull in (excluding colab).

[pinhodecarlos]

Originally Posted by KEP

Maybe some of the users active at SETI@home, should refer the guys there and their attention to putting their GPU to good use, now that SETI@home stops feeding their GPUs after march 31st. If all the SETI@home users moved their GPUs to srbase to help GIMPs, that would really blow the ceiling and send us beyond the Moon - compared to the firepower we already see.

Have you begun spreading the word about GPU72?

Only on my team forum but Linux client is still not stable. Problem is that the majority of SETI GPU users are now running COVID-19 on another project called GPUGrid. Let’s just hope a tiny percentage joins this effort in the future. Fingers crossed.

Let me give you a small example of this COVID impact on Boinc. Syracuse University, top NFS@Home user, moved all its cpu power to run covid-19. Its daily output is equivalent to 4,250 of my home laptop.

GPUGrid current GigaFLOPs 2,727,600. As soon as the app levar it’s alpha stage people will receive feed news on boinc about this new app.

Edit: I haven’t yet posted on Gridcoin forum about this.

[KEP]

Originally Posted by pinhodecarlos

Only on my team forum but Linux client is still not stable. Problem is that the majority of SETI GPU users are now running COVID-19 on another project called GPUGrid. Let’s just hope a tiny percentage joins this effort in the future. Fingers crossed.

Let me give you a small example of this COVID impact on Boinc. Syracuse University, top NFS@Home user, moved all its cpu power to run covid-19. Its daily output is equivalent to 4,250 of my home laptop.

GPUGrid current GigaFLOPs 2,727,600. As soon as the app levar it’s alpha stage people will receive feed news on boinc about this new app.

Okay, again this Corona thing is hampering our excistense and what could be done and achieved. Lets hope that once this usefull use of GPUs to fight Corona wears off, that just a few percent of those users will move to srbase. Maybe as you mention, they come all by themself, once we go out of testing mode

[pinhodecarlos]

Originally Posted by KEP

Okay, again this Corona thing is hampering our excistense and what could be done and achieved. Lets hope that once this usefull use of GPUs to fight Corona wears off, that just a few percent of those users will move to srbase. Maybe as you mention, they come all by themself, once we go out of testing mode

What I advice is when the client leaves the testing phase to trigger the feed news, not sure how BOINC admins do that. On the boinc client there’s a section for it so I can easily let you guys know when it is received.