Folding Forum

What's the difference between Timeout and Expiration date?

See http://folding.stanford.edu/English/FAQ-main#ntoc24

Its clear now.

Thank you.

I forgot that I can look-up in the FAQ.

your reply is page not found. yay progress.

Read here: https://foldingathome.stanford.edu/supp ... ldinghome/

tldr for anyone stumbling across this thread (for if/when the link breaks again)

• unfinished work units “expire” and are reassigned to new machines.
  • You will still receive credit for all WUs completed and uploaded prior to the Timeout
  • Even so, full credit is given up until the Deadline
  • After the Deadline has expired, the client will automatically discard the work unit
• after the Timeout,
  • your contribution is not as useful scientifically because another copy of that work unit had to be sent out to another contributor.
  • Even if you eventually complete the work unit, that other contributor still had to process duplicate work to assure the science moves forward.
  • If you have trouble completing work units before the Timeout…, it is recommended to either run the FAH client more hours each day, or to run the client on a faster computer.

(the only reason I'm bothering is because this thread is quite high up on relevant Google queries)

Hi,

thanks for summing this up but I think I still don't get it.

I see two dates for each WU in FAHControl. "Timeout" and "Expiration". "Timeout" is usually about 1 day after assignment, "Expiration" about 1 week after assignment.

Did I get this right, that after the "Timeout" deadline (1 day after assignment) my contribution is "not as useful scientifically because another copy of that work unit had to be sent out to another contributor."? But I still get credit if I upload the WU before the "Expiration" deadline (1 week after assignment)?

If so, I'm in trouble, because I use FAH on a relatively slow computer (Intel Atom), but it runs 24/7 and it's always struggeling to finish a WU before the first "Timeout" deadline and missed it yet some times.

Greetings

Marginauto wrote:Did I get this right, that after the "Timeout" deadline (1 day after assignment) my contribution is "not as useful scientifically because another copy of that work unit had to be sent out to another contributor."? But I still get credit if I upload the WU before the "Expiration" deadline (1 week after assignment)?

Correct, except for one detail:
The WU is not necessarily given out again immediately, but it becomes possible from that point on (I'd say it depends on the number of clients requesting WUs vs. the availability of other WUs).

You are not in trouble!

But you may not be contributing to the science needed to combat diseases.

It would save you some electric bill to quit folding, and your WU may go to someone else to fold a day sooner.

Thanks for your explanations.

With "being in trouble" I meant "Having spent some electricity for the last weeks for FAH and having enyoyed watching my user statistics without being aware that most of my WUs were useless". And, yes, of course I contribute to FAH because I want to help the science and not to get some shiny virtual gamifications points.

Are there some statistics in FAHControl, which can show me how many WUs were finished in time (in definition of "Timeout" and not "Expiration")?

I don't now how many people beside me have these problems, but I'd strongly suggest to equalize the two deadlines to avoid people folding useless. Some warnings in FAHControl like "You already missed the deadline for the 3rd WU. We'd suggest you to stop folding on this computer" were great too.
Another way would be to create smaller WUs for slower computers, so that they can also contribute to FAH.

When F@H has an excess of WUs to fold, you might have still been the first to finish, as other volunteers got other WUs. But with the current shortage of WUs, I bet 'your' WU gets reassigned promptly.

It is true that there are both CPUs and GPUs that can meet the minimum specifications to fold but cannot meet the deadlines. My slowest 2.2 Ghz T6600 Core2 Duo finishes about 3 hours before timeout 24/7

It is possible your Intel Atom could support a modern GPU so it could fold.

I would not say "useless", but perhaps of less scientific value in some cases.

JimboPalmer does overstate the case a bit. For example, if you got a WU, completed it in 25-26 hours, then it would probably be the first returned and then creates the next generation of that trajectory to be assigned. Very unlikely the WU would have been reassigned and returned in less than a few hours. The researchers will still give that "insurance" reassignment whatever points it earns, and sometimes they do compare multiple returns to see if they are consistent.

Now as the researchers are getting away from the initial rush of getting projects out and wanting very fast returns, more of the CPU projects are getting longer Timeout values than 1 day. So meeting the Timeout may become less of an issue for you.

As for getting Bonus, i.e. returning WUs before Timeout, if you enter the username that you are using here - https://apps.foldingathome.org/cpu, that shows the results for the last WU folded on each of your systems. Making the assumption that you use the same name as you have registered with here, that shows the last 4 WU returns all got bonus.

Marginauto wrote:Did I get this right, that after the "Timeout" deadline (1 day after assignment) my contribution is "not as useful scientifically because another copy of that work unit had to be sent out to another contributor."? But I still get credit if I upload the WU before the "Expiration" deadline (1 week after assignment)?

The deadlines of 1 day and 7 days are what I consider unusual. They started appearing when the COVID19 projects created an inrush of new donors. Recently I'm seeing a lot more projects with 2 and 5, In fact, the timeout is supposed to be long enough that midrange GPUs or CPUs can finish a good percentage of the WUs, making any duplicate that might get issued the one that may become a wasted effort.

The atom is a low-performance CPU but don't count it out. They do complete their share of WUs. Remember you can't predict what might happen to the other half of the duplication process when it does happen.

Marginauto wrote:Thanks for your explanations.

With "being in trouble" I meant "Having spent some electricity for the last weeks for FAH and having enyoyed watching my user statistics without being aware that most of my WUs were useless". And, yes, of course I contribute to FAH because I want to help the science and not to get some shiny virtual gamifications points.

Are there some statistics in FAHControl, which can show me how many WUs were finished in time (in definition of "Timeout" and not "Expiration")?

I don't now how many people beside me have these problems, but I'd strongly suggest to equalize the two deadlines to avoid people folding useless. Some warnings in FAHControl like "You already missed the deadline for the 3rd WU. We'd suggest you to stop folding on this computer" were great too.
Another way would be to create smaller WUs for slower computers, so that they can also contribute to FAH.

You can view your logs and compare points awarded vs base points that display at https://apps.foldingathome.org/psummary

If you are awarded base points or above that amount, you met the timeout or turned it in before the timeout.

There are also ways to check specific WU's for return points, or use software that monitors the time of the WU's you run. Using HFM allows you to compare against the timeout times easily.


As for your previous scenario of your system barely meeting or exceeding the timeout, points are still awarded after that due to the work involved. Let's say your system takes 30 hours to complete a WU that has a timeout of 1 day (24 hours). Yes that WU can be assigned to another folder, but there is no guarantee that the folder that gets it can complete it in 6 hours. That folder might be able to do it in 2, or it might take more than 30. So they award points as an incentive to attempt to keep folding and turn in the needed science first.

In the above scenario, if the system that picks up your WU is 5 times faster than yours, their result might be turned in first. And there are such systems, be it CPU or GPU folding. But what if another system on par with yours picks it up? It would be another 30 hours starting from scratch vs your 6 more hours to complete the one you had running.

Hi,

thank you for your explanations. I'll keep on folding and will have an eye on the statistics.

Greetings