I do calculations on a laptop. In order to reduce noise, my computer does very few calculations. Often it does not have time to finish the calculations of your project before the deadline. At the moment, you send tasks of 80000 GFLOPs. It would be very convenient if you had tasks of 15000 GFLOPs. The deadline for this should be increased from 3 days to 5. Thank you!

I do calculations on a laptop. In order to reduce noise, my computer does very few calculations. Often it does not have time to finish the calculations of your project before the deadline. At the moment, you send tasks of 80000 GFLOPs. It would be very convenient if you had tasks of 15000 GFLOPs. The deadline for this should be increased from 3 days to 5. Thank you!

It's entirely up to the project to decide how soon tasks need to be returned to them, not users.
It's a bit churlish to reduce the resources you put into the project to suit you, then complain that the project's tasks take too long - including for everyone else.

It seems you have a decently fast machine - faster than mine for example - so you could either put back some of that resource
Or you could reduce your target runtime by going to your online settings, entering rosetta@home preferences and changing "Target CPU run time" to a shorter time than the default 8hrs
Some combination of those two settings that you're comfortable with should allow you to meet deadlines without asking the project to change things project-wide to suit you alone.

---

And the usual issue for not getting work done in time....

Run time 4 hours 37 min 19 sec
CPU time 3 hours 11 min 18 sec

4.5 hours to do 3 hours of work.

Set "Use at most xx % of CPU time" to 100%, and "Use at most xx % of the CPUs" to something along the lines of 20% (or less if necessary).
The system will have no issues doing the default Target CPU time setting, and it will have no issues with the deadlines if you set your cache accordingly- given that you run more than one project there is no need for a cache to keep the system busy, so 0.1 days and 0.01 additional days would be plenty for your cache. And there will be no issues with noise from the fan running for all it's worth.

---

Grant
Darwin NT

And the usual issue for not getting work done in time....

Run time 4 hours 37 min 19 sec
CPU time 3 hours 11 min 18 sec

4.5 hours to do 3 hours of work.

Set "Use at most xx % of CPU time" to 100%, and "Use at most xx % of the CPUs" to something along the lines of 20% (or less if necessary).
The system will have no issues doing the default Target CPU time setting, and it will have no issues with the deadlines if you set your cache accordingly- given that you run more than one project there is no need for a cache to keep the system busy, so 0.1 days and 0.01 additional days would be plenty for your cache. And there will be no issues with noise from the fan running for all it's worth.

And that runtime over 2 days 15hrs
Only 1 Rosetta task seems to come down at a time, so I suspect % of CPUs is currently set at 5% of 20 cores - lol
All of which was wasted because the tasks are getting aborted because they're miles off completion.
Asking % of CPU time to be set at 100% might be a bit hopeful. It should at least be increased by, say, an extra 10% then reduce target CPU runtime to be 4hrs to be successful, I'm guessing, and to avoid the manual intervention of aborting tasks.

As paltry as that is, users are entitled to decide for themselves what they want to contribute (in this case without noisy fan intervention) but the minimum needs to allow success, not repeated intervention and failure, otherwise they're wasting all the time they're running with no-one seeing any benefit. That can't serve anyone's purpose.

Edit: I've just glanced at the other projects being run and they're barely any more successful than Rosetta, with multiple tasks having to be aborted. No-one is winning here.
Increasing % of CPU cores is only going to increase the number of tasks having to be aborted, so definitely don't do that.
Increasing % of CPU time will help all projects to be successful - the more without CPU fan intervention the better - so after an initial extra 10%, if there's scope for any more it'll only serve to reduce those interventions further, which I can well imagine must be annoying. My 2 cents
And yes, keeping the CPU cache size to the bare minimum will allow tasks to start running very quickly after download, so is probably worth doing as well. Every bit will help.

---