Seti@Home optimized science apps and information

Optimized Seti@Home apps => Discussion Forum => Topic started by: Mongo on 31 Aug 2008, 06:26:48 pm

Title: AK app for SSE3 and SSE4.1 returned different spike count
Post by: Mongo on 31 Aug 2008, 06:26:48 pm

While browsing my completed SETI@Home tasks, I found a fellow SAH user's Q6600 Quad machine, using Alex's SSE3 v8.0 app returned 2 spikes for the same WU that my E8400 machine, using the SSE4.1 v8.0 app, reported 0 spikes.    Both machines are running Windows XP Pro, his on SP3, mine still on SP2.

Q6600 result is here:
http://setiathome.berkeley.edu/result.php?resultid=969767281

E8400 result is here:
http://setiathome.berkeley.edu/result.php?resultid=969767282

Shouldn't the counts be the same?

Mongo

Title: Re: AK app for SSE3 and SSE4.1 returned different spike count
Post by: Jason G on 31 Aug 2008, 09:48:15 pm

Hi there,

 The key here is in the "Restarted at 89.15 percent.".  This means the very much cosmetic signal counts (they do not constitute any part of the result stored at the science database) reset after the resume, and continue counting.  We may may an adjustment in a future build to include that in checkpointing but have not considered it a priority.

Jason

Title: Re: AK app for SSE3 and SSE4.1 returned different spike count
Post by: Mongo on 31 Aug 2008, 09:52:39 pm

Quote from: Jason G on 31 Aug 2008, 09:48:15 pm

Hi there,

 The key here is in the "Restarted at 89.15 percent.".  This means the very much cosmetic signal counts (they do not constitute any part of the result stored at the science database) reset after the resume, and continue counting.  We may may an adjustment in a future build to include that in checkpointing but have not considered it a priority.

Jason

Thanks!  Appreciate the clarification.

[/worries]

Mongo