point disparity and inconsistency on an alpha project

Message boards : Projects : point disparity and inconsistency on an alpha project
Message board moderation

To post messages, you must log in.

AuthorMessage
Profile Joseph Stateson
Volunteer tester
Avatar

Send message
Joined: 27 Jun 08
Posts: 551
United States
Message 21433 - Posted: 21 Nov 2008, 15:27:47 UTC

I recently signed up with milkyway, an alpha project, and noticed some pecularities that I have not yet gotten an answer for.

Should not a workunit detail show the same granted credit at the top (summary) as at any of the detail lines at the bottom. For example http://tinyurl.com/65ruup shows 260.00 granted credits at the top but 96.69 at the bottom. Compare that to the ABC home which matches correctly http://tinyurl.com/65ruup

In addition, this kruncher has a really old 366mhz celeron (x86 family 6 model 6) and only ran for 3,222 seconds, just under an hour to get either 260 or 96.69 depending on what you believe.

What also seems inconsistent is that, a 366mhz Pentium 3 (celeron) gets more credit in less than an hour than a Q6700 or Q9750 core gets on ABC in an hour and a half or more. Something does not seem right.
I saved a copy of the workunit detail in the event this one "disappears" by the time you try to read it as their database has run out of room and most results are being deleted in less than an hour.
http://swri.info/cpustats/mw_celeron_366.png

There is a thread about the above here
http://tinyurl.com/5u4jr6
ID: 21433 · Report as offensive
Nicolas

Send message
Joined: 19 Jan 07
Posts: 1174
Argentina
Message 21435 - Posted: 21 Nov 2008, 17:21:48 UTC - in response to Message 21433.  

Heh, *every* stats site forum (and MilkyWay itself of course) has a full flamewar (er, I mean "thread") about this :)
ID: 21435 · Report as offensive
mo.v
Avatar

Send message
Joined: 13 Aug 06
Posts: 778
United Kingdom
Message 21466 - Posted: 23 Nov 2008, 19:16:50 UTC
Last modified: 23 Nov 2008, 19:43:07 UTC

Thank you, Dagorath. I knew you would stand up for crunchers' rights.

On CPDN Beta we had to wait about 4 months for our credits, then only some of us got them. So they were taken away and the other people got theirs instead. So those credits were taken away and the first group got theirs back until they were again taken away. After waiting another month we all got our credits except a few members who'd used the wrong email address. Only two people complained and one of those complaints was really a polite query.

If on MilkyWay you get 96 credits for under 1 hour of crunching on a slowish computer, this is still 4x what we get per hour on CPDN or CPDN Beta with a fairly fast machine. And we have to back our climate models up and restore them if they crash, sometimes losing days of crunching. So MilkyWay crunchers should feel very, very lucky and should not start flame wars.
ID: 21466 · Report as offensive

Message boards : Projects : point disparity and inconsistency on an alpha project

Copyright © 2021 University of California. Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation.