| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Community
Forum: Chat Room
Thread: DRAM Errors in the Wild: A Large-Scale Field Study |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 5
|
|
| Author |
|
|
TimAndHedy
Senior Cruncher Joined: Jan 27, 2009 Post Count: 267 Status: Offline Project Badges: 
|
I just noticed this thread over at XtremeSystems.
----------------------------------------It seems like interesting information especially in relation to WCG. http://www.xtremesystems.org/forums/showthread.php?t=236252 Several of the conclusions were interesting but conclusion 3 includes "Given that DRAM DIMMs are devices without any me- chanical components, unlike for example hard drives, we see a surprisingly strong and early effect of age on error rates. For all DIMM types we studied, aging in the form of in- creased CE rates sets in after only 10-18 months in the field." Something the developers will need to keep in mind especially considering the number of non-ecc systems. [Edit 1 times, last edit by TimAndHedy at Oct 9, 2009 2:22:24 AM] |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
http://www.tech-faq.com/ecc-memory.shtml
tells you what ecc memory does sa far as i can tell its %5-20 slower but doesnt make the mistakes normal ram does couldnt work out the% nice to know we are a certain % wrong |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
Was there an imply here that some crunched results due 'random DIMM error' might have undetected flaws? If so, think not it has any relevance given all the other verification procedures before even getting close to a lab test.
----------------------------------------
WCG
----------------------------------------Please help to make the Forums an enjoyable experience for All! [Edit 1 times, last edit by Sekerob at Oct 15, 2009 11:29:29 AM] |
||
|
|
TimAndHedy
Senior Cruncher Joined: Jan 27, 2009 Post Count: 267 Status: Offline Project Badges:
|
Two purposes.
1. For people to understand their systems could fail randomly and potentially why. My systems run 24 hours a day. I would expect WCG systems run more that average in general. Age is also a factor and it seems like there are many older systems running here. 2. Yes, the potential to effect results did cross my mind. Projects like HCC that require two identical results obviously are immune. What about projects like HFCC that don't require two results. What are these processes for verification? I scanned the faq but did not see anything. |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
Those ZR jobs are grouped in the 10's of thousands, same as HPF2 and the gravitation towards best energy results / best dockings, most common and real life matching folds are sought out. So as commented by Dr. Watowich, first pass, second pass, in hopes to be left with the 20 or so best fitting compounds or whatever objective the scientists have set for their project.
----------------------------------------
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
|