Quantcast

Maximum PC

It is currently Tue Sep 16, 2014 8:45 pm

All times are UTC - 8 hours




Post new topic Reply to topic  [ 4 posts ] 
Author Message
 Post subject: Folding Core A4 seems to have a bug in it :(
PostPosted: Sun Nov 04, 2012 11:02 am 
Team Member Top 500
Team Member Top 500
User avatar

Joined: Thu Jul 21, 2005 10:45 pm
Posts: 3026
Location: Central Florida
It's not a security issue, so no worries there, but there seems to be a bug in the A4 (SMP) core that causes a tiny fraction of wu's to run with very slow frame times. It may also cause the work unit to give incorrect data, so even though you get credited for the completion, the data may be useless to Stanford.

This post at the Stanford forums has some info on the subject: including this post from Bruce
bruce wrote:
Both Kasson and tjlane have seen WUs in their project which were completed after running with a high TPF and determined that the result was corrupt. They have examples demonstrating that there is a bug in one or more of the SMP cores and the core developer is aware of the issue but it has not yet been fixed.

It's not appropriate to assume that this bug is the ONLY reason someone might have a WU that has a high TPF. That's one reason why I asked if the step count changed. It could help them to understand if it's the same problem or a different one.


This bug has apparently been around since the beginning of the A4 core, but seems that some work units, especially the 600 point p70xx work units, may be causing it to show up a bit more than it had in the past.

Here are the two I've experienced, but there are others.
http://foldingforum.org/viewtopic.php?f=19&t=22869
http://foldingforum.org/viewtopic.php?f=19&t=22821

If you find any of these please post the results over at the Stanford Forums. This bug has so far been near the bottom of the 'to fix' list, if it's becomming a more wide-spread problem, they need to know so it can be elevated to a higher urgency level. I'm sure they evenutally see the issue as they're compiling the completed WU's, but a heads up gives them something specific to look for & to find right away.


Top
  Profile  
 
 Post subject: Re: Folding Core A4 seems to have a bug in it :(
PostPosted: Sun Nov 04, 2012 8:20 pm 
Team Member Top 10
Team Member Top 10

Joined: Sat Jul 09, 2005 7:50 am
Posts: 1427
Thanks for the info on the possible bug in the A4 core wu. I'll be keeping a watch for high frame times and especially with the A4 and report it in this thread and at Stanford forum. I'm just wondering whether to finish the wus after and if it's obviously of the suspected bugged p70xx unit? If we can get them noticed more at Stanford, then hopefully they can better make an informed decision to rewrite some of it, or get lucky and pop that bug where it is... splat! Nice going, chaosdsm. :wink:


Top
  Profile  
 
 Post subject: Re: Folding Core A4 seems to have a bug in it :(
PostPosted: Mon Nov 05, 2012 6:43 am 
I fear no heat
I fear no heat

Joined: Fri Mar 25, 2005 7:39 am
Posts: 299
Location: MA Folding as David_W
How do I check if this bug affects me?

My PPD took a huge hit over the weekend even though I'm still folding full force.
http://folding.extremeoverclocking.com/ ... =&u=124340

Image


Top
  Profile  
 
 Post subject: Re: Folding Core A4 seems to have a bug in it :(
PostPosted: Mon Nov 05, 2012 8:02 am 
Team Member Top 500
Team Member Top 500
User avatar

Joined: Thu Jul 21, 2005 10:45 pm
Posts: 3026
Location: Central Florida
The official stance is: IF you can verify that it is the bug & not something else causing the slow-down, then report your findings at the Stanford forum, dump the work unit, change machine ID (otherwise you're likely to get the exact same WU again), & restart folding. This is because, even though the WU completes, & shows as no errors, the data is in fact corrupt & of no use to Stanford.

Dav5049915 wrote:
How do I check if this bug affects me?

My PPD took a huge hit over the weekend even though I'm still folding full force.
http://folding.extremeoverclocking.com/ ... =&u=124340

Image

Holy cow, that's a nasty looking drop :shock:

I check progress several times a day, and am very familiar with what work units "should" be producing on my machine. The bug in question still leaves the CPU running at or near 100% utilization, but frame completion times are drastically slowed. For example, I know that the 600 point work units should run at about 1min 45sec to 2min 10sec between frames, but the bug causes the frame rates to jump up to about 8min to 9min per frame on my machine.

If you're running the V7 client, you can select either Advanced or Expert views from the drop down menu on the top right corner. This will bring up an extra box on the display that shows current work unit details including time per frame. If the bug were to show up on a work unit like the one below, it could have a dramatic affect on your production, especially if it multiplies the time per frame on every work unit it affects.
Image

You can also use HFM to monitor & track work unit progress & history.


Top
  Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 4 posts ] 

All times are UTC - 8 hours


Who is online

Users browsing this forum: No registered users and 3 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group