Welcome to MilkyWay@home

Questions on Nbody

Message boards : Application Code Discussion : Questions on Nbody
Message board moderation

To post messages, you must log in.

AuthorMessage
Kylinblue

Send message
Joined: 23 Aug 11
Posts: 7
Credit: 498,188
RAC: 0
Message 64371 - Posted: 10 Mar 2016, 19:17:40 UTC
Last modified: 10 Mar 2016, 19:18:16 UTC

Run time seems varies by a lot. I checked computation size and they were similar, but wall time varies too much, one can be finish in 10 minutes and the other took 10+ hours. Another thing I noticed is that when I run Nbody task, the handles in task manager skyrockets, right now it is at 1 million...


ID: 64371 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Jake Weiss
Volunteer moderator
Project developer
Project tester
Project scientist

Send message
Joined: 25 Feb 13
Posts: 580
Credit: 94,200,158
RAC: 0
Message 64806 - Posted: 5 Jul 2016, 13:32:11 UTC

Kylinblue,

The run time varies a lot because the step size in the simulation depends on the parameters going into the simulation. This means for some parameter combinations, there are fewer steps than other parameter combinations. This leads to varying simulations times. Sidd should be trying to keep the simulations from hitting excessively large simulation times, but it seems he didn't do a good enough job for this simulation.

I will talk to Sidd and remind him to be nice to you guys and keep the simulations to reasonable times.

Jake
ID: 64806 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Thunder
Avatar

Send message
Joined: 9 Jul 08
Posts: 85
Credit: 44,842,651
RAC: 0
Message 64814 - Posted: 5 Jul 2016, 16:59:01 UTC

It's worth noting that you do get more credit for the longer N-Body tasks.

How credit is computed is way too complicated to get into here, but suffice to say that you're getting roughly the same amount of credit per amount of time, regardless of how long the tasks take.

For me, in this last round of N-Body, I've seen tasks as short as about 2 minutes and as long as 170 hours (on the same machine, I might add) so as long as yours are falling inside that, it's probably all good. :-)
ID: 64814 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Curtis Owens

Send message
Joined: 30 Oct 13
Posts: 2
Credit: 230,994
RAC: 360
Message 70714 - Posted: 4 Apr 2021, 12:04:47 UTC - in response to Message 64814.  

This seems to still be an issue. Is there no way to combine these tests so that the fast ones and slow ones are known? Or even better combined so that 2 tests or enough tests are combined to average out the CPU usage and the expected completion time.
ID: 70714 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote

Message boards : Application Code Discussion : Questions on Nbody

©2024 Astroinformatics Group