rpi_logo
new searches testing different validation types
new searches testing different validation types
log in

Advanced search

Message boards : News : new searches testing different validation types

Author Message
Profile Travis
Volunteer moderator
Project administrator
Project developer
Project tester
Project scientist
Send message
Joined: 30 Aug 07
Posts: 2046
Credit: 26,480
RAC: 0

Message 35909 - Posted: 22 Jan 2010, 4:08:11 UTC

I've implemented optimistic and pessimistic validation for the particle swarm searches now as well (as opposed to just the differential evolution searches that were running before). I've tested them quite a bit so I think it's working correctly, but just warning you the server might crash a few times in the next few days while I get all the bugs sorted out.

Let me know if you're having any problems with the ps_s222_opt_1, ps_s222_pes_1 workunits.
____________

Profile The Gas Giant
Avatar
Send message
Joined: 24 Dec 07
Posts: 1947
Credit: 240,884,648
RAC: 0

Message 35918 - Posted: 22 Jan 2010, 9:34:49 UTC

On my 4870 I'm seeing about a 13% increase in crunch time (for the same credit) and what's interesting on my 4850 I'm only seeing about a 2% increase in crunch time.

Profile arkayn
Avatar
Send message
Joined: 14 Feb 09
Posts: 999
Credit: 74,932,619
RAC: 0

Message 35922 - Posted: 22 Jan 2010, 12:24:47 UTC - in response to Message 35918.

My 4830 is still running the same times from what I can tell.

304 to 308 seconds.
____________

Raimund Barbeln
Send message
Joined: 7 Oct 07
Posts: 25
Credit: 29,063,967
RAC: 0

Message 35927 - Posted: 22 Jan 2010, 14:52:35 UTC

My 4770s are taking about 30% longer for the pes WUs, the opt WUs take the same time as the old WUs.
____________
When life gives you lemons, make lemonade!

Profile Travis
Volunteer moderator
Project administrator
Project developer
Project tester
Project scientist
Send message
Joined: 30 Aug 07
Posts: 2046
Credit: 26,480
RAC: 0

Message 35934 - Posted: 22 Jan 2010, 20:16:29 UTC - in response to Message 35918.

On my 4870 I'm seeing about a 13% increase in crunch time (for the same credit) and what's interesting on my 4850 I'm only seeing about a 2% increase in crunch time.


There really shouldn't be any difference in compute time. The only changes were server-side.
____________

Profile The Gas Giant
Avatar
Send message
Joined: 24 Dec 07
Posts: 1947
Credit: 240,884,648
RAC: 0

Message 35942 - Posted: 22 Jan 2010, 22:32:36 UTC
Last modified: 22 Jan 2010, 22:36:26 UTC

Times were definitely longer initially, but on looking at them now they have gone back to 'normal'.

[edit] Sort of, the last one on my 4870 was 18 seconds longer (10%), so I'm seeing them move around a bit.


Post to thread

Message boards : News : new searches testing different validation types


Main page · Your account · Message boards


Copyright © 2018 AstroInformatics Group