| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
| World Community Grid Forums
|
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 15
|
|
| Author |
|
|
Paul van Dijken
Cruncher Joined: Aug 15, 2011 Post Count: 5 Status: Offline Project Badges:
|
I recently got a few hundred MCM tasks. They are estimated to run in 1.45 hours, but they need 7 hours to complete. With a short deadline (only a few days), many of them have not started before the deadline expires and they get aborted by BOINC.
I am not running anything else. Anyone else have similar problems? |
||
|
|
alanb1951
Veteran Cruncher Joined: Jan 20, 2006 Post Count: 1316 Status: Recently Active Project Badges:
|
Do the running tasks appear to be clocking up progress properly? If not, have you got "Keep applications in memory" set for when things suspend? If applications get swapped out if suspended (e.g. because of your "when computer in use" settings) then they'll restart at the last checkpoint...
----------------------------------------If that's not the issue, a little insight into what your system is (O/S, CPU version, BOINC client version?) might help, but here are some things that could be relevant whatever those details might be.
The other points might have some bearing as well, because although it may appear to be running things all the time, if you have a large number of tasks running at the same time it will tend to mean that each clock cycle sees less actual computing because of either floating point delays or higher volumes of memory accesses that can't be satisfied from cache. I don't know whether the above is of any help, but... Cheers - Al. [Edited to add the reference to keeping applications in memory!] [Edit 2 times, last edit by alanb1951 at Nov 7, 2025 11:02:34 AM] |
||
|
|
Garrulus glandarius
Advanced Cruncher Romania Joined: Apr 10, 2025 Post Count: 88 Status: Offline Project Badges:
|
Points 1 and 2 can definitely lead to massive increases in runtime. I have a laptop with an Ultra 5 125H CPU that has a mix of performance and efficiency cores. Performance cores take about 2x as long to finish similar tasks. I also disabled "turbo" to avoid running the CPU at 100C (now it runs at 60-70C @ 2 GHz) so I ended up with very long runtimes.
----------------------------------------In comparison, a low-powered Ryzen 7 5825U throttled to 80C (resulting in ~3-3.1 GHz) but used as a dedicated BOINC rig performs much much better. I recently added a 8845HS to my BOINC farm, also throttled it to 80C (runs at around 4 GHz) and it has a 50% higher output than the 5825U. Unsurprisingly, it also uses almost 50% more power. ![]() ![]() |
||
|
|
Paul Schlaffer
Senior Cruncher USA Joined: Jun 12, 2005 Post Count: 278 Status: Offline Project Badges:
|
Is an additional cache set up at the bottom of the Computing Preferences menu? If so, eliminate it.
----------------------------------------BOINC will adjust the estimated time to the actual after multiple WU are processed, and use that number for fetching work in the future.
“Where an excess of power prevails, property of no sort is duly respected. No man is safe in his opinions, his person, his faculties, or his possessions.” – James Madison (1792)
|
||
|
|
Garrulus glandarius
Advanced Cruncher Romania Joined: Apr 10, 2025 Post Count: 88 Status: Offline Project Badges:
|
Is an additional cache set up at the bottom of the Computing Preferences menu? If so, eliminate it. BOINC will adjust the estimated time to the actual after multiple WU are processed, and use that number for fetching work in the future. I have 0,3 + 0. I set both to 0 now, see how it goes. Funny that my best CPU is estimated to run tasks 5x slower than it actually does. ![]() ![]() |
||
|
|
Paul van Dijken
Cruncher Joined: Aug 15, 2011 Post Count: 5 Status: Offline Project Badges:
|
Thank you.
The tasks appear to be clocking fine. I have checked the keep applications in memory (btw, they are not swapped out, they run continuously) OS: Win11, Boinc client: Latest CPUs: 16 cores, 3056 floating point MIPS, 17479 integer MIPS 11th Gen Intel core i7 11800H @ 2.3GHz Use % is set to 100%, but is actually 8%. Clockspeed is reduced to 1 GHz (and my laptop is getting warm) |
||
|
|
alanb1951
Veteran Cruncher Joined: Jan 20, 2006 Post Count: 1316 Status: Recently Active Project Badges:
|
Paul van Dijken,
Try reducing number of cores available to BOINC to (say) 8 and see what difference that makes. Thermal throttling is probably taking place at present (and you'd probably return more tasks a day with 8 "CPUs" than with (say) 12; the sweet spot may be even lower... My small fleet is on a restricted energy budget (old house, limited total power availability!) so I try to keep most systems running below their notional peak power using CPU frequency capping -- I sometimes see thermal throttling if the ambient temperature passes 25C (where 18..20C is "normal"), but not very often even then. For what it's worth, I once killed a Skylake-based laptop by not keeping an eye on temperatures -- it didn't seem to be running hot but it eventually started giving audible alarms and by then it was too late! Fortunately, modern CPUs will throttle back on overheating :-) And on another occasion a brand new Ryzen 3700 suffered a CPU fan failure and until that was noticed it was running one CPU thread at 0.2 GHz with impressively high run times as a result (which was how I spotted it...) -- as I said, modern CPUs are better protected :-) Good luck getting it sorted out... Cheers - Al. |
||
|
|
alanb1951
Veteran Cruncher Joined: Jan 20, 2006 Post Count: 1316 Status: Recently Active Project Badges:
|
Garrulus glandarius,
In reverse order (and drifting off the original topic, I fear)... Funny that my best CPU is estimated to run tasks 5x slower than it actually does. The initial estimates of time are based on the [absolutely useless] BOINC benchmarks and it can take quite a while for the system to work out what is right. If that's on a P-core, E-core system that could also confuse things!...a low-powered Ryzen 7 5825U throttled to 80C (resulting in ~3-3.1 GHz) but used as a dedicated BOINC rig performs much much better. Two of my (all-Linux) fleet members are mini-PCs; one is an Asus 5600H which I frequency cap at 3.6GHz and the other a MinisForum 7840HS which I cap at 4.1GHz. I also moderate the work mix (never allow more than 75% of CPU threads to be used by BOINC) and I divvy up the possible use of those threads by application; both systems have very respectable turnaround times, even when under a full workload -- on a good day the 5600H returns up to 120 MCM1 tasks a day (limited to 5 threads) and the 7840HS returns over 200 (limited to 8 threads) whilst only drawing about 25% more power. Nice little systems, and there's definitely a performance-per-watt improvement with the latest Zen architecture!I recently added a 8845HS to my BOINC farm, also throttled it to 80C (runs at around 4 GHz) and it has a 50% higher output than the 5825U. Unsurprisingly, it also uses almost 50% more power. Cheers - Al. |
||
|
|
Garrulus glandarius
Advanced Cruncher Romania Joined: Apr 10, 2025 Post Count: 88 Status: Offline Project Badges:
|
The initial estimates of time are based on the [absolutely useless] BOINC benchmarks and it can take quite a while for the system to work out what is right. If that's on a P-core, E-core system that could also confuse things! The Ryzens only have P-cores, completely identical ones. SiDock has been running on the 8845HS for several days non-stop, but the estimates remain the same (well over 2 days when in reality they run 9-10 hours). Two of my (all-Linux) fleet members are mini-PCs; one is an Asus 5600H which I frequency cap at 3.6GHz and the other a MinisForum 7840HS which I cap at 4.1GHz. I also moderate the work mix (never allow more than 75% of CPU threads to be used by BOINC) and I divvy up the possible use of those threads by application; both systems have very respectable turnaround times, even when under a full workload -- on a good day the 5600H returns up to 120 MCM1 tasks a day (limited to 5 threads) and the 7840HS returns over 200 (limited to 8 threads) whilst only drawing about 25% more power. Nice little systems, and there's definitely a performance-per-watt improvement with the latest Zen architecture! Nice! I also recently became a fan of miniPCs dedicated to BOINC. Smaller, cheaper, much quiter and using much less power than full-sized rigs and way cheaper than similar-powered laptops. The issue is finding enough space for them and also not getting carried away with the shopping spree. I'm running the minis 24/7, all cores on BOINC. ![]() ![]() |
||
|
|
|