Advanced search

Message boards : Number crunching : Work Units????

Author Message
CodeRedDewd
Send message
Joined: 11 Nov 09
Posts: 27
Credit: 4,925,174
RAC: 0
Level
Ala
Scientific publications
watwatwatwatwat
Message 48291 - Posted: 7 Dec 2017 | 2:28:44 UTC

I use my PC as a heater during the winter and just started running again since it will be 45 tonight in Phoenix. Are there no work units to be downloaded? It keeps saying "communications deferred" and it says there is one work unit ready to download. Is there not enough work???

PappaLitto
Send message
Joined: 21 Mar 16
Posts: 511
Credit: 4,672,242,755
RAC: 0
Level
Arg
Scientific publications
watwatwatwatwatwatwatwat
Message 48292 - Posted: 7 Dec 2017 | 13:11:58 UTC - in response to Message 48291.

I use my PC as a heater during the winter and just started running again since it will be 45 tonight in Phoenix. Are there no work units to be downloaded? It keeps saying "communications deferred" and it says there is one work unit ready to download. Is there not enough work???

Currently not, I suggested expand the scientists that can contribute to this supercomputer as rushing scientists to get out work faster is not going to get anything done.

wiyosaya
Send message
Joined: 22 Nov 09
Posts: 114
Credit: 589,114,683
RAC: 0
Level
Lys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 48294 - Posted: 7 Dec 2017 | 18:10:57 UTC - in response to Message 48292.

I use my PC as a heater during the winter and just started running again since it will be 45 tonight in Phoenix. Are there no work units to be downloaded? It keeps saying "communications deferred" and it says there is one work unit ready to download. Is there not enough work???

Currently not, I suggested expand the scientists that can contribute to this supercomputer as rushing scientists to get out work faster is not going to get anything done.

Personally, I find it amazing that these NO WORK WTF? posts continually pop up. So please excuse the following rant.

For a user like us that donates computing power, it is easy to say to the project should "add whatever at whatever cost". However, the project has limited budget. If they had an unlimited budget, they would buy their own supercomputer and would not need us to supply computing power.

The research field, such as what is done by this project - research - is rarely highly profitable. It is expensive to perform research, and this project has a donations button because they would love to have the extra money. Even if someone donates, it is a donation and the project will allocate any donation to what they determine will be the best place to spend that donation.

Most research projects have a limited budget and can only do so much. I imagine that all BOINC projects have a very long wish list, and since they have a limited budget, they have prioritized their wish lists and will fill that wish list as they deem necessary and when the items are within their budget and priority.

It is not the responsibility of any BOINC project to provide work units to any user for any reason. Not in the least. There are plenty of other GPU projects out there that have work to do such as Einstein, MilkyWay, Amicable Numbers, and Seti to name a few.

Please take my advice: Join those projects and remember that we are the volunteers here; we volunteer our computing power to run the work units of BOINC projects. Not all projects will have work units at all times for any number of reasons whether we like it or not. If that is not acceptable, then I suggest not running BOINC.

Again, please excuse my rant.

Jim1348
Send message
Joined: 28 Jul 12
Posts: 819
Credit: 1,591,285,971
RAC: 0
Level
His
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 48297 - Posted: 7 Dec 2017 | 20:51:20 UTC - in response to Message 48294.


It is not the responsibility of any BOINC project to provide work units to any user for any reason. Not in the least. There are plenty of other GPU projects out there that have work to do such as Einstein, MilkyWay, Amicable Numbers, and Seti to name a few.

Please take my advice: Join those projects and remember that we are the volunteers here; we volunteer our computing power to run the work units of BOINC projects. Not all projects will have work units at all times for any number of reasons whether we like it or not. If that is not acceptable, then I suggest not running BOINC.

Again, please excuse my rant.

I agree completely. If they had all the money they wanted, they wouldn't need us.

CodeRedDewd
Send message
Joined: 11 Nov 09
Posts: 27
Credit: 4,925,174
RAC: 0
Level
Ala
Scientific publications
watwatwatwatwat
Message 48301 - Posted: 8 Dec 2017 | 3:40:59 UTC - in response to Message 48294.

This is the only one I've asked this about, because it's a better project with better purpose than using my GPUs to find out if 1=0 in some strange case. That's the only reason I asked.

Betting Slip
Send message
Joined: 5 Jan 09
Posts: 670
Credit: 2,498,095,550
RAC: 0
Level
Phe
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 48304 - Posted: 8 Dec 2017 | 11:29:55 UTC - in response to Message 48301.

This is the only one I've asked this about, because it's a better project with better purpose than using my GPUs to find out if 1=0 in some strange case. That's the only reason I asked.


I can't agree with you that this is a "better project" with a "better purpose".

All well run projects have the potential to make discoveries that are for the 'good' or 'ill' of mankind whether it be Maths or Biology or mapping Asteroids. Never forget that some of the most groundbreaking discoveries in science have been made by accident while looking for something else.

A solution to a math problem can be just as important to the human race as a solution to a biological one and they're also both double edged swords as regard to what mankind uses them for.

The reason Bio projects get more support is because as humans we see them as a doorway to perfect health and eventually immortality so it's very easy to relate to.

Profile JStateson
Avatar
Send message
Joined: 31 Oct 08
Posts: 186
Credit: 3,361,527,550
RAC: 283,434
Level
Arg
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 48305 - Posted: 8 Dec 2017 | 16:41:56 UTC
Last modified: 8 Dec 2017 | 16:43:12 UTC

if this lasts through the upcoming crunchathlon, I can see all teams tied for first place at 0 units each!

Profile Retvari Zoltan
Avatar
Send message
Joined: 20 Jan 09
Posts: 2343
Credit: 16,201,255,749
RAC: 0
Level
Trp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 48310 - Posted: 8 Dec 2017 | 21:14:34 UTC - in response to Message 48291.

I use my PC as a heater during the winter and just started running again since it will be 45 tonight in Phoenix.
Me too. It is wise then to set up backup projects (=zero resource share set on the given project's computing preferences).

Are there no work units to be downloaded? It keeps saying "communications deferred" and it says there is one work unit ready to download. Is there not enough work???
It is intentional that there's no work in the queue, as this project needs very short turnaround times because the workunits are generated from the result of a previous workunit, so the queue is kept as short as possible (=0). There are always around 2200-2500 workunits in progress, which means that the staff continuously fills the queue with work, but it is instantly grabbed by some hosts.

mmonnin
Send message
Joined: 2 Jul 16
Posts: 337
Credit: 7,198,776,065
RAC: 13,733,598
Level
Tyr
Scientific publications
watwatwatwatwat
Message 48314 - Posted: 10 Dec 2017 | 0:36:51 UTC

I also have a backup project with E@H set to 0%. Some days my two GPUs do 1/4 to 1/2 work for E@H and not for GPUGrid.

Profile JStateson
Avatar
Send message
Joined: 31 Oct 08
Posts: 186
Credit: 3,361,527,550
RAC: 283,434
Level
Arg
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 48325 - Posted: 10 Dec 2017 | 15:11:07 UTC - in response to Message 48314.

I also have a backup project with E@H set to 0%. Some days my two GPUs do 1/4 to 1/2 work for E@H and not for GPUGrid.


I have never messed with resource share before, how does this help?

I have one system with gtx 1070 TI and another with the slower card 1060. Occasionally the slower one has an active gpugrid WU and another queued ready to run while the faster 1070Ti has not seen a gpugrid WU in days. Both systems run e@h with 3 apps on the 1060 and 4 on the 1070 as memory is 3 and 8(4) respectivel

Another question: 1070 has 8gb but opencl shows only 4 available. I assume GPU projects can only use the 4gb.

Betting Slip
Send message
Joined: 5 Jan 09
Posts: 670
Credit: 2,498,095,550
RAC: 0
Level
Phe
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 48328 - Posted: 10 Dec 2017 | 15:21:04 UTC - in response to Message 48325.



Another question: 1070 has 8gb but opencl shows only 4 available. I assume GPU projects can only use the 4gb.


GpuGrid App is 32bit so can only see 4GB Max

mmonnin
Send message
Joined: 2 Jul 16
Posts: 337
Credit: 7,198,776,065
RAC: 13,733,598
Level
Tyr
Scientific publications
watwatwatwatwat
Message 48341 - Posted: 12 Dec 2017 | 21:05:47 UTC - in response to Message 48325.

I also have a backup project with E@H set to 0%. Some days my two GPUs do 1/4 to 1/2 work for E@H and not for GPUGrid.


I have never messed with resource share before, how does this help?

I have one system with gtx 1070 TI and another with the slower card 1060. Occasionally the slower one has an active gpugrid WU and another queued ready to run while the faster 1070Ti has not seen a gpugrid WU in days. Both systems run e@h with 3 apps on the 1060 and 4 on the 1070 as memory is 3 and 8(4) respectivel

Another question: 1070 has 8gb but opencl shows only 4 available. I assume GPU projects can only use the 4gb.


With the resource share at 0% for a project it will not download any work until your processor is out of work and you have idle resources. There will be no queued work for that project. You'll only get what you can immediately work on.

So if you have run out of GPUGrid work it will download a single E@H task. If while you're working on that task you receive a GPUGrid task, E@H will not download any more work until a GPU is idle.

I try to leave a backup project for each GPU and CPU I have. Typically a project that has pretty stable work like E@H for GPUs and WCG for CPUs.

Matt Falcon
Send message
Joined: 4 Nov 17
Posts: 9
Credit: 40,025,925
RAC: 0
Level
Val
Scientific publications
watwat
Message 48488 - Posted: 24 Dec 2017 | 5:23:42 UTC

I'll add that I'm in the same boat - I have a small studio and use 3 computers as heaters - it's >100% efficient! ;) Instead of just converting electricity to heat, it converts electricity to computing and then, in turn, into heat! Outside temp = 37F, inside temp = 67F.

Longer rationale short, I've noticed lately that I haven't had any GPU work for my 1080 or 660Ti. Been checking every morning and at least for the past 2 weeks, they've been crunching SETI instead of GPUGrid.

My GPUGrid priority is set to 1500 to out-prioritize others set to 100 (SETI) and 50 (E@H). But it still ends up downloading tons and tons and tons of SETI WUs...

I've also got a profile set up for the 660Ti computer to only download "short work" units for GPUGrid. Turns out, it crunches through those in the same time the 1080 takes to run a long WU, so it was pretty nice!

On the 1080, since it's under-worked by SETI, I also have Folding@Home running at the same time. But since the two platforms operate independently, I have to back-off F@H when it finds a GPUGrid WU. Kind of a bugger to juggle, but that's why I hope for a steady stream of work I can have run on autopilot...

Really just posting this as a me-too and a case-in-point describing my setup. Some tips would be nice but I'm not complaining :)

Erich56
Send message
Joined: 1 Jan 15
Posts: 1120
Credit: 8,967,195,176
RAC: 31,162,806
Level
Tyr
Scientific publications
watwatwatwatwatwatwatwatwat
Message 48489 - Posted: 24 Dec 2017 | 6:39:03 UTC - in response to Message 48488.
Last modified: 24 Dec 2017 | 6:39:47 UTC

I've also got a profile set up for the 660Ti computer to only download "short work" units for GPUGrid.

there have not been any short runs for quite some time.
No idea when new ones will be made available.

mmonnin
Send message
Joined: 2 Jul 16
Posts: 337
Credit: 7,198,776,065
RAC: 13,733,598
Level
Tyr
Scientific publications
watwatwatwatwat
Message 48496 - Posted: 24 Dec 2017 | 13:50:04 UTC - in response to Message 48488.

I'll add that I'm in the same boat - I have a small studio and use 3 computers as heaters - it's >100% efficient! ;) Instead of just converting electricity to heat, it converts electricity to computing and then, in turn, into heat! Outside temp = 37F, inside temp = 67F.

Longer rationale short, I've noticed lately that I haven't had any GPU work for my 1080 or 660Ti. Been checking every morning and at least for the past 2 weeks, they've been crunching SETI instead of GPUGrid.

My GPUGrid priority is set to 1500 to out-prioritize others set to 100 (SETI) and 50 (E@H). But it still ends up downloading tons and tons and tons of SETI WUs...

I've also got a profile set up for the 660Ti computer to only download "short work" units for GPUGrid. Turns out, it crunches through those in the same time the 1080 takes to run a long WU, so it was pretty nice!

On the 1080, since it's under-worked by SETI, I also have Folding@Home running at the same time. But since the two platforms operate independently, I have to back-off F@H when it finds a GPUGrid WU. Kind of a bugger to juggle, but that's why I hope for a steady stream of work I can have run on autopilot...

Really just posting this as a me-too and a case-in-point describing my setup. Some tips would be nice but I'm not complaining :)


I have my other projects, also E@H and S@H, set to 0% so they carry no extra queue. When there are GPUGrid tasks you have a full queue and are less likely to get tasks. To entire my GPUs are busy during downloads I have them running 2x tasks.

mikey
Send message
Joined: 2 Jan 09
Posts: 294
Credit: 4,853,873,615
RAC: 22,168,100
Level
Arg
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 48517 - Posted: 28 Dec 2017 | 2:56:41 UTC - in response to Message 48496.

I'll add that I'm in the same boat - I have a small studio and use 3 computers as heaters - it's >100% efficient! ;) Instead of just converting electricity to heat, it converts electricity to computing and then, in turn, into heat! Outside temp = 37F, inside temp = 67F.

Longer rationale short, I've noticed lately that I haven't had any GPU work for my 1080 or 660Ti. Been checking every morning and at least for the past 2 weeks, they've been crunching SETI instead of GPUGrid.

My GPUGrid priority is set to 1500 to out-prioritize others set to 100 (SETI) and 50 (E@H). But it still ends up downloading tons and tons and tons of SETI WUs...

I've also got a profile set up for the 660Ti computer to only download "short work" units for GPUGrid. Turns out, it crunches through those in the same time the 1080 takes to run a long WU, so it was pretty nice!

On the 1080, since it's under-worked by SETI, I also have Folding@Home running at the same time. But since the two platforms operate independently, I have to back-off F@H when it finds a GPUGrid WU. Kind of a bugger to juggle, but that's why I hope for a steady stream of work I can have run on autopilot...

Really just posting this as a me-too and a case-in-point describing my setup. Some tips would be nice but I'm not complaining :)


I have my other projects, also E@H and S@H, set to 0% so they carry no extra queue. When there are GPUGrid tasks you have a full queue and are less likely to get tasks. To entire my GPUs are busy during downloads I have them running 2x tasks.


I do the same thing but would love it if GpuGrid would have more workunits on a consistent basis.

Post to thread

Message boards : Number crunching : Work Units????

//