Too large WU/ too short deadline /no way to refuse

[Oleg F]

Hi, guys, seriously!

WU
16990
My computer has 4.04 days to complete this work unit.
Estimated time 1.39 days, and efficiency seems to be floating, at the beginning work time expected was about 4 days. 4 days of work of 5 days deadline.
I can't spend 80% of working time, and FAH system used to accept task automatically, so that neither I can indicate working time, nor I can refuse the work due to no intention do it till end.

That is I'm either forced to lend my PC work 96h on full, or just fail common effort.
It's kinda totalitarian. Please don't turn into totalitarian governement, believe me, it's not cool, totally.
So I just forced to stop task and disable CPU slot to avoid useless energy wasting.

Somebody Do Something! =)

[JimboPalmer]

Yes, a new Work Unit does take some percentage to stabilize a predicted completion and therefore estimated points.

Dumping a WU is never good for F@H, so they do not make it easy.

16990 is a Core_a8 project for CPUs, F@H is generally well behaved on CPUs, running at low priority and letting you use the PC without slowdown. (If your PC is low on memory, that can slow it down, swapping memory)

[aetch]

Project 16990 has a timeout of 4 days and an expiry of 4.2 days (about 4 days and 5 hours).

Ignore the initial estimate ETA of the work unit, when the client receives a work unit for a new project it uses the timeout as the ETA.
Only once it it has a few folds logged (3%-4%) can it better estimate the ETA, so 1.39 days sounds a bit more reasonable.

Most of us run our system(s) 24/7, many of us have dialled them down so they're less power hungry and quieter because the cooling fans don't spin as fast.
Not everybody is happy to run 24/7. Something I've seen other suggest is to reduce the number of cpu cores/threads available to your cpu slot to 4 so you will see smaller projects (a number of projects have a minimum cpu/thread count threshold). The caveat to that is that if you reduce it too far your processor will take longer to process the work unit, which defeats the purpose of reducing it in the first place.

Also bear in mind that folding is not for everyone so don't worry if you cannot contribute.

[Oleg F]

Yes, it says 1.11 days now, so, probably, shorter ETA is closer to reality. Probably long work was misleading numbers and me.
It says CPU:6, so seems like it uses 6 threads on 8 virtual cores.
Thank you for answers!

Still, question of "I want skip particular WU", i.e. quit doing it and immediately inform system the WU have to be redistributed is actual.
First of all, it's question of free will.
And there are cases like:
a) person going to go away from particular PC for some time(week etc.) but got WU too big.
b) not enough time left to finish job, like WU to take 1 day, 2 days left and person won't be running PC in idle 50% of time.
c) it's Summer 2020, we have 30 Celsiuses of hell in our rooms here, burning GPU is really difficult, pricey electricity, rampaging coronavirus and I get the job with desc "Hello, World! It's our first work unit on coronavirus blah-blah-blah", come on, guys, half of America already died out of this nightmare virus, we must hurry saving the world, so I just didn't want to do particular one, deleted unit somehow.

So, skipping WU voluntarily, indicating working time owner prefer and can afford to give out are questions, really.

[JimboPalmer]

Dumping WUs is very bad for F@H, if you are tempted frequently, it is better not to fold at all. Once a year, not so bad.

[aetch]

I'm going to assume you're using Windows 10

Changing the number of cpu cores/threads
Start -> Folding@Home -> FAHControl
FAHControl -> Configure (button top left)
Configure -> Slots (3rd tab) -> Select cpu type slot (usually at top of Folding Slots frame) -> Edit
Folding Slot -> CPUs change number as required (by default this is set to -1, this allows the FAHClient to control thread allocation, each GPU gets a thread and the CPU slot gets the remainder), I'd suggest setting it to 4.

Lowering CPU speed/power draw/heat generation
Start -> Windows System -> Control Panel
Control Panel-> View By: Large/Small Icons (Menu, Top Right) -> Power Options -> Change plan settings -> Change advanced power settings
Power Options (new window) -> Change settings that are currently unavailable (this may or may not appear) -> From dropdown select power plan which has "[Active]" marked beside it ->
Processor Power management -> Maximum Power State -> Plugged in: <change as required, 100% allows cpu to turbo/boost if supported, 99% and lower disables turbo/boost and lowers the cpu frequency> (the cpu speed won't be as granular as this control panel setting)

Lowering GPU speed/power draw/heat generation
Download and install MSI Afterburner
Run MSI Afterburner -> Find the "Power Limit (%)" silder, slide to left -> Click the big apply tick button (it's under the "Fan Speed (%)" slider)

[josh_rodman]

Found this thread and it feels applicable because I have a 24/7 Ryzen 3700x rig with an (?) 8-core/12 hyperthreaded slot that got assigned a CPU unit for project #18423 that initially estimated at 14 days ( !!!!!!! ) and even now, two (24-hour) days later, it estimates completion in 12 days. That 12 days is both beyond both the timeout AND the expiration, despite running every day all day.

There is simply no point to wasting energy on this unit in my rig. I have to believe there was some kind of mishap or this project's WU completion time was poorly estimated or somehow meant for a different type of system. I've never seen this kind of WU size before and obviously sending it to my rig is only delaying useful F@H submissions.

In any case .. I'd like to fold on that slot but don't know how to dump the unit on the Windows 10 client. I've found instructions for Linux but that's not what I'm running so they aren't terribly helpful. Anyone know how to dump a unit on Windows 10 .. particularly if there's a way to signal the server that this unit CANNOT complete before the expiration (and really .. shouldn't the software after like 20% processing proactively kill and reassign a unit that cannot be completed on that slot even in a 24/7 time period?) I've got two GPU running full bore in that folding rig and those units are clicking off no problem so uninstalling/reinstalling isn't a very appealing option (and it doesn't signal F@H to reassign that unit).

Any help is appreciated!!

Note: Project #18423 (5, 1, 9)

[gunnarre]

You could pause the CPU slot until the WU expires, perhaps, but let the GPU slots run?

[Joe_H]

@ josh_rodman I suspect you have something not set up well on your system, either overcommitted CPU threads or a background process taking up CPU time. Normally your system should be capable of completing these WUs well within the time limits.

During beta testing I did work on some WUs from this series of projects. My system has an non-overclocked i7-7700K that I use 4 of the available 8 CPU threads, it took about 25 hours to complete a WU. That is well within the 7 day timeout for bonus points and the 10 day expiration deadline.

My first suggestion is to reduce the thread count to the number of native CPU cores - 8. More threads than that require the FPU to be shared between threads, that can result in limited gains in processing time.

The second suggestion is if you do not want to get these [rojects assigned at all is to set your CPU thread count to 4. That currently is the cutoff, only systems with more than 4 threads will get these assigned. A value of greater than 4 was picked due to the number of dual core systems with HT out there requesting for 4 threads.

[nemric]

Same problem here, my CPUs are smal (ATOM D525 and D2700) and recently my FAHclient download unit I can't achieve on 24/7 running computers, ETA is calculated over timeout date but this cause no problem to the client, it runs, for nothing...
I asked for downloading smaller WU but the best would be to adjust timeouts on CPU capabilities or giving arbitrary 2 or 3 more days...
I know these are very small CPUs but they are for fan-less/noise-less home servers and the purpose of FAH is to use free cpu time not to force users to buy big noisy machines to fold 24/7
for now, project 16969 and 16955 have a deadline really too short for me... (My 2 atom CPUs at least, the celeron G3220T without hyperthreading works great)
at this time, 21-12-01 , the D525 announce ETA 5.98 days, timeout 21-12-03 12:04, expiration date 21-12-05

[JimboPalmer]

Let me be brutal to be honest.

A WU that is not returned or returned too late not only slows your WU, but it slows the entire project. The next generation cannot start until all of this generation is finished.

As such, there is no incentive for the developer to cater to slow CPUs. She/he want results as quick as they can get them. "Publish or Perish" is very real in Academia.

https://boinc.berkeley.edu/
I can and do offer help to speed up marginal CPUs, but at some point, it is time to find a BOINC project where speed is not important.

[nemric]

Yes you're brutal !

I fold since 2004, I give free cpu time (and in some way, it's money) to this project for years !

The time I fold too large units is lost for science due to timeout limit that is too short, so don't assign so much work to small CPU ... If science don't need CPU's free time, it's time to say it now !

There are many posts here that talk about that, for PCs that are shutdown during the night, during weekend, and so on ... A lot of work is lost due to timeout and you claim that science need us !? science don't have time to wait for ongoing WUs but have the time to reassign the work that have been done up to 90% by a slow machine ? perhaps it will be assigned to a poor cpu again and again to have 90% + 90% .... of the work done but never published ?

You'd better optimize your process before asking me to upgrade my PC to work faster ! Science should not care about who have the biggest one !

"Publish or Perish" !? I have 3 PCs that work 24/7 for fah ... are they useless ? do you talk for science or for yourself ?

[v00d00]

The time I fold too large units is lost for science due to timeout limit that is too short, so don't assign so much work to small CPU ... If science don't need CPU's free time, it's time to say it now !

"Publish or Perish" !? I have 3 PCs that work 24/7 for fah ... are they useless ? do you talk for science or for yourself ?

If you aren't making deadlines then it probably is time to deactivate those machines. It's just a matter of fact with FAH. Over time machines get old and if you don't upgrade they eventually don't meet deadlines and you are then damaging the project. Usually the project maintainers get around this by creating cores that no longer work on older hardware or won't use older optimisations. But users are expected to act responsibly and decommision when the time is right. CPU folding is a dying art nowadays unless you are fielding big systems with 32+ cores. You can generally do more for science with a cheap nvidia card that uses very little watts than with an old dual core cpu.

So it's up to you in the end. You can continue running something that is damaging the science of the project due to selfishness or just accept that the project moves forward and scales up with upgrades in technology, which in time, means older technology becomes obsolete. But the responsibility is yours, not ours.

[nemric]

[JimboPalmer]

Note: Virtually everyone here giving help is a volunteer. Very rarely, a volunteer with pull will ask for server help from F@H staff.

I have retired at least 6 graphics cards and over a dozen CPUs since volunteering, my oldest CPU is a second generation i3, my oldest graphics card a GTX1060. It happens.