16/04/2020 22:28:32 | Einstein@Home | see scheduler log messages on https://einsteinathome.org/host/11977460/log

So, what it needs?? OpenCL here... what app requirements then?

is asking for 1 second of CUDA work, but says

[HOST#11977460] is looking for work from a non-preferred application

Have you ticked 'Use NVidia GPU' in project preferences?

https://clip2net.com/s/4746ZFq
Seems too old GPU for modern E@h indeed :/

---

SETI apps news
We're not gonna fight them. We're gonna transcend them.

Perhaps we should all set up a new team called Seti Orphans, and go invade another project.. ? :-)

I like such idea! Maybe on different projects but the same team, SETI Orphans :)
Would be the honour to participate !
P.S. my personal choice would be Einstein@home as closest in spirit one.
Will see if it can feed all my different devices. Not compatible ones will search for new targets...

. . It would be an honour to follow :)

Stephen

:)

Well, I went to F@h and found no SETI Orphans team there !

So I created one!

Team number: 262364

Welcome to join !
(they promise to give COVID19 related work for GPUs )

EDIT: but seems it would be not so easy to make it coexist with BOINC.
It already took 1 CPU core... but left GPU idle :/

And no possibility to select COVID19, only cancer, parkinson and so on.

---

SETI apps news
We're not gonna fight them. We're gonna transcend them.

The trick to managing work at E@H is to set a really, really small cache. Like 0.1days. Then you basically run on a "turn one in - get one" schedule.

. . I set their resource share to 0% and get one for one without changing cache settings.

Stephen

:)

About E@h:

I installed old NV 460SE GPU from another (currently disabled) host, but Einstein refuses to use it:

16/04/2020 22:28:15 | | CUDA: NVIDIA GPU 0: GeForce GTX 460 SE (driver version 365.19, CUDA version 8.0, compute capability 2.1, 1024MB, 950MB available, 746 GFLOPS peak)
16/04/2020 22:28:15 | | OpenCL: NVIDIA GPU 0: GeForce GTX 460 SE (driver version 365.19, device version OpenCL 1.1 CUDA, 1024MB, 950MB available, 746 GFLOPS peak)
and
16/04/2020 22:28:18 | Einstein@Home | Sending scheduler request: To fetch work.
16/04/2020 22:28:18 | Einstein@Home | Requesting new tasks for NVIDIA GPU
16/04/2020 22:28:32 | Einstein@Home | Scheduler request completed: got 0 new tasks
16/04/2020 22:28:32 | Einstein@Home | No work sent
16/04/2020 22:28:32 | Einstein@Home | see scheduler log messages on https://einsteinathome.org/host/11977460/log
16/04/2020 22:28:32 | Einstein@Home | Gamma-ray pulsar binary search #1 on GPUs is not available for your type of computer.
16/04/2020 22:28:32 | Einstein@Home | Gravitational Wave search O2 Multi-Directional is not available for your type of computer.
16/04/2020 22:28:37 | Rosetta@home | Sending scheduler request: To fetch work.
16/04/2020 22:28:37 | Rosetta@home | Requesting new tasks for NVIDIA GPU
16/04/2020 22:28:39 | Rosetta@home | Scheduler request completed: got 0 new tasks


So, what it needs?? OpenCL here... what app requirements then?

I think your card falls short in the available memory department. I found this old post at Einstein about minimum gpu requirements.

Right now I think the current GPU-linux requirements are:

- OpenCL 1.1 capable graphics card (this means GeForce 4xx and above)

- minimum 1 gig of memory on the graphics card

- 64 bit linux

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

16/04/2020 22:28:32 | Einstein@Home | see scheduler log messages on https://einsteinathome.org/host/11977460/log

So, what it needs?? OpenCL here... what app requirements then?

is asking for 1 second of CUDA work, but says

[HOST#11977460] is looking for work from a non-preferred application

Have you ticked 'Use NVidia GPU' in project preferences?

Hi,
Einstein seems to have open source code and it uses 45-88% GPU on my system.
Seems like it could need some optimizing. But but,... if it is OpenCL then I'd have to turn it to CUDA first ...

Hi petri, keep in mind they have two different types of tasks there, the low GPU utilized tasks must be the Gravity Wave task. it requires a lot of spare CPU support. I've seen 150% CPU thread per GPU. The Gamma Ray tasks pull only 100% CPU thread and have good 98% GPU utilization.
I would love to see if optimizations are possible there! tilt the scales back towards the nvidia cards :). as it stands, the project heavily favors AMD cards.

. . Hey, if this happens I might turn my 2 * GTX970 and 2 * GTX1060 rigs back on :) Those rigs have enough CPU threads to support both GPUs.

:)

Hi,
Einstein seems to have open source code and it uses 45-88% GPU on my system.
Seems like it could need some optimizing. But but,... if it is OpenCL then I'd have to turn it to CUDA first ...

. . Stay tuned folks, this could get interesting :)

Stephen

:)

But I do not feel the connection with either project (however worthwhile) that I have enjoyed with S@H.

Stephen

:(

+1

---

"I owe", "I owe", "Its off to work I go" (from a bumper sticker on a smallish Mercedes Benz)
(on the back of a Semi Tractor) "If you can read this bumper sticker, I've LOST MY TRAILER!"

Einstein will do the same if you give it a small resource share, because it has a 7 day deadline. It insists on filling up to your cache size, then doesn't run for 5 days or so then panics and has to do more work than the resource share allows. Then come the Seti outrage and it would refill the cache, wash, rinse, repeat.

. . I am getting 14 day deadlines at E@H, 7 days I think at WCG and 72 hours at Rosetta. But as a backup I ran/run E@H with 0% resource share and I get 1 task at time so I have one waiting to report and one running. This works pretty well for me.

Stephen

:)

About E@h:

I installed old NV 460SE GPU from another (currently disabled) host, but Einstein refuses to use it:

16/04/2020 22:28:15 | | CUDA: NVIDIA GPU 0: GeForce GTX 460 SE (driver version 365.19, CUDA version 8.0, compute capability 2.1, 1024MB, 950MB available, 746 GFLOPS peak)
16/04/2020 22:28:15 | | OpenCL: NVIDIA GPU 0: GeForce GTX 460 SE (driver version 365.19, device version OpenCL 1.1 CUDA, 1024MB, 950MB available, 746 GFLOPS peak)
and
16/04/2020 22:28:18 | Einstein@Home | Sending scheduler request: To fetch work.
16/04/2020 22:28:18 | Einstein@Home | Requesting new tasks for NVIDIA GPU
16/04/2020 22:28:32 | Einstein@Home | Scheduler request completed: got 0 new tasks
16/04/2020 22:28:32 | Einstein@Home | No work sent
16/04/2020 22:28:32 | Einstein@Home | see scheduler log messages on https://einsteinathome.org/host/11977460/log
16/04/2020 22:28:32 | Einstein@Home | Gamma-ray pulsar binary search #1 on GPUs is not available for your type of computer.
16/04/2020 22:28:32 | Einstein@Home | Gravitational Wave search O2 Multi-Directional is not available for your type of computer.
16/04/2020 22:28:37 | Rosetta@home | Sending scheduler request: To fetch work.
16/04/2020 22:28:37 | Rosetta@home | Requesting new tasks for NVIDIA GPU
16/04/2020 22:28:39 | Rosetta@home | Scheduler request completed: got 0 new tasks


So, what it needs?? OpenCL here... what app requirements then?

---

SETI apps news
We're not gonna fight them. We're gonna transcend them.

Personally I was just pondering what I was going to do for this years WOW event. Whether in fact I would enter at all as it struck me last year as not very environmentally friendly.
Seems the decision has been taken out of my hands. I will crunch on my 4 main machines until there is nothing left to crunch, then I will be left with 5 750 Ti's, 2 1050's, 2 1060, and a 970 with nothing to do.
Whether I will crunch for another project remains to be seen, SETI was really the only one that ever interested me.

. . I thought you had a 1660ti as well? But I feel much like you. I have left three machines on E@H because it was my 'backup' project but I only do their tasks on the GPUs. To utilise the CPU I have tried Rosetta but had several issues there so I moved to WCG where I am not experiencing problems. But I do not feel the connection with either project (however worthwhile) that I have enjoyed with S@H.

Stephen

:(

Open one "SETI orphans" Café or Pub in every country city where we have SETI@home participants :-)

Stephen

:)

I'll go to a project that has an open source code for Nvidia GPUS and an off-line test suite similar to Seti@home.

. . Hey there Petri, nice to see you here.

. . Let us know if/where you set up shop, I for one will probably join you :)

Stephen

..

I've been wondering what effect this news will have on the rest of the scientific community?

Will they ...
... see lots of skilled users with powerful computers, ready to take on a new area of science?
... mis-understand and read it (wrongly) as the beginning of the end of the BOINC platform as a whole?

. .Many are running out of work repeatedly because they are unaccustomed to such large numbers of volunteers :)

Stephen

:)

Happened to me a couple of times. I shortened the run times to 4 hours so that I was less likely to lose any time. I'm only at rosetta for the covid, once that project is done I'm trying another project.

Same here. I will assign non-24/7 hosts to another "location" with shortest projected run time possible.
And stable "might" hosts will get longest possible projected run time to reduce start new task overheads.

---

SETI apps news
We're not gonna fight them. We're gonna transcend them.

Even if Rosetta were to explicitly state their applications are not suitable for Mobile phones, people would still use them & complain. Or complain that they don't have an application they can use on their phone...

Haha, yes definitely will! :D :D :D
,

---

SETI apps news
We're not gonna fight them. We're gonna transcend them.

Happened to me a couple of times. I shortened the run times to 4 hours so that I was less likely to lose any time. I'm only at rosetta for the covid, once that project is done I'm trying another project.

WCG will start a COVID-19 project also pretty soon. It's in Alpha testing now, for more than a week, and is expected to go Beta any day now.
Please feel free to join through this link:

World Community Grid

And if you do, also consider joining our Team on WCG "Seti Orphans"

3 day deadlines?
At least WCG gives one 7-10 days. :-)

The big problem is not just that 3 days, is the lack of stop points on some WU.
If for any reason you need to stop boinc (totally exit not just suspend) and restart you could loose hours of crunching for nothing.
That happening to me 2 days ago. I have a Roseta 4.15 running for >8 hrs (with about 80% done) and due a system update i was forced to reboot the host.
In my ignorance i take the usual steps. Totally exit boinc, and wait until it gives the ok i'm down. Then stop all services on the host and do the usual cold boot.
When everything was working again for my surprise, the task restart from 2 hrs approximately (at about 25%) loosing more than 6 hrs of crunching.
Looking on the Rosetta forum, there are a lot complains about this.

Happened to me a couple of times. I shortened the run times to 4 hours so that I was less likely to lose any time. I'm only at rosetta for the covid, once that project is done I'm trying another project.

3 day deadlines?
At least WCG gives one 7-10 days. :-)

The big problem is not just that 3 days, is the lack of stop points on some WU.
If for any reason you need to stop boinc (totally exit not just suspend) and restart you could loose hours of crunching for nothing.
That happening to me 2 days ago. I have a Roseta 4.15 running for >8 hrs (with about 80% done) and due a system update i was forced to reboot the host.
In my ignorance i take the usual steps. Totally exit boinc, and wait until it gives the ok i'm down. Then stop all services on the host and do the usual cold boot.
When everything was working again for my surprise, the task restart from 2 hrs approximately (at about 25%) loosing more than 6 hrs of crunching.
Looking on the Rosetta forum, there are a lot complains about this.

---