i'm seeing very long run time for my tasks on BOINC and i've been told that it's because something is eating up a lot of my CPU time, but i'm looking at my task manager and i don't see any other apps taking up CPU time other than the tasks. i wonder if there's anything else i can tweak or check. thanks!

BOINC version: 7.24.1 (x64)
OS: Windows 10
Projects: WCG, E@H, R@H

i'm seeing very long run time for my tasks on BOINC and i've been told that it's because something is eating up a lot of my CPU time, but i'm looking at my task manager and i don't see any other apps taking up CPU time other than the tasks. i wonder if there's anything else i can tweak or check. thanks!

BOINC version: 7.24.1 (x64)
OS: Windows 10
Projects: WCG, E@H, R@H

What is a very long run time. Back in the day, CPDN used to have tasks that would last 6 months on a slow machine!
Go to the project websites and check how long others are taking for tasks from the same batches. A bit of hunting will confirm for certain whether they are really taking a long time or not. Also worth checking you are not short on RAM as swapping out to disk will slow some things down. (more likely I suspect with Einstein than WCG. How many threads are you using. (My main work CPDN slows down noticeably if I use more than the 8 real cores. Maximum throughput is N-1 tasks where N is the number of real cores in my experience. Also worth checking nothing in your computing preferences is slowing things down. Difficult to give any other hints without a bit more information.

Be sure that the task manager is showing you all tasks and not just the tasks for your user.

It’s very often the background system tasks that steal your resources - windows updates, file indexer, McAfee, etc.

It’s very often the background system tasks that steal your resources - windows updates, file indexer, McAfee, etc.

Thanks. So used to Linux and typing, "top" in a terminal that shows everything I didn't think of that.

It’s very often the background system tasks that steal your resources - windows updates, file indexer, McAfee, etc.

Thanks. So used to Linux and typing, "top" in a terminal that shows everything I didn't think of that.

I think that this was issue in WIndows 7 but Windows 10 does not have that selection anymore. It shows all tasks by default.

by long run time i mean:

Run time 19 hours 40 min 22 sec
CPU time 2 hours 59 min 49 sec

this is unusual as someone else running R@H said theirs is more like:
Run time 3 hours 3 min 58 sec
CPU time 3 hours 0 min 6 sec

as for computing preferences, mine is:
Usage limits
Use at most 100 % of the CPUs
Use at most 100 % of CPU time

When to suspend
Suspend when computer is on battery leave blank
Suspend when computer is in use leave blank
Suspend GPU computing when computer is in use leave blank
'In use' means mouse/keyboard input in last 3 minutes
Suspend when no mouse/keyboard input in last --- minutes
Suspend when non-BOINC CPU usage is above --- %
Compute only between ---

Other
Store at least 0.1 days of work
Store up to an additional 0.01 days of work
Switch between tasks every 60 minutes
Request tasks to checkpoint at most every 60 seconds

Disk
Use no more than 20 GB
Leave at least 2 GB free
Use no more than 60 % of total

Memory
When computer is in use, use at most 95 %
When computer is not in use, use at most 95 %

i don't know how to check how many threads i'm using or how to set that as a preferred mode. i also don't know what else anyone might need to know to help...

i don't know how to check how many threads i'm using or how to set that as a preferred mode. i also don't know what else anyone might need to know to help...

You have BOINC set to use 100% of CPUs. If you had the same CPU as my main machine, that would mean 16 tasks running on a CPU with 8 real cores. So with my machine I set it to use 49% of CPUs which cut it down to 7 cores running. If you look in the fora of the individual projects you run it is likely somewhere there is discussion about the optimum number of cores to use for those projects. With WCG for instance the number of cores which produces maximum throughput is higher with MCM tasks than it was with ARP ones when they were available.

by long run time i mean:

Run time 19 hours 40 min 22 sec
CPU time 2 hours 59 min 49 sec

this is unusual as someone else running R@H said theirs is more like:
Run time 3 hours 3 min 58 sec
CPU time 3 hours 0 min 6 sec

This depends highly on hardware capabilities. You haven't said anything about what your CPU (and possibly GPU) is.

(My main work CPDN slows down noticeably if I use more than the 8 real cores. Maximum throughput is N-1 tasks where N is the number of real cores in my experience. Also worth checking nothing in your computing preferences is slowing things down. Difficult to give any other hints without a bit more information.

I am running tasks for 3 LHC projects, and one each on Einstein and Rosetta. Two of the LHC projects (Atlas and CMS) are multi-CPU tasks. My CPU is a 12-core/24 thread Ryzen 9 5900X. Currently, I have told Boinc to use 22 of those 24 threads.
Are you suggesting that I would see greater throughput if I set Boinc to use only 11 or 12 CPUs instead of the 22? I fail to understand how that could be true. Each core implements hyperthreading, so is capable of running 2 tasks nearly at full speed. The biggest bottleneck would be when both tasks require the math unit at the same time -- as there is only one math unit per each core.
Plus, if I told Boinc to use only half the CPUs, then I could no longer run Atlas and CMS on multiple threads, and still be able to run a few tasks from the other 3 projects.

Are you suggesting that I would see greater throughput if I set Boinc to use only 11 or 12 CPUs instead of the 22? I fail to understand how that could be true. Each core implements hyperthreading, so is capable of running 2 tasks nearly at full speed. The biggest bottleneck would be when both tasks require the math unit at the same time -- as there is only one math unit per each core.

I have no idea where the sweet spot of maximum throughput is with projects other than CPDN as I haven't done the tests. I would be surprised if no one at LHC has done them though. My understanding from others on the CPDN boards is that it is to do with the level3 cache getting hammered by the CPDN tasks. My machine is starting to get a little old running with PC3200 DDR4 ram. I don't know how the goalposts change on this with faster RAM.