Gunsvnvmc
Posting anything negative at all on the Stanford Forum gets removed and then you get a nasty gram from them. So how do we get there attention and help fixing the problems on there end. They have the habit of saying everything is the fault of everyone but themselves.

Shut systems down for a month. Force them to pay attention.

I have posted on their forums now, and I will wait to see their response.

Here's the problem as I see it - Stanford has to try to craft a solution based on the best tools that are available - Currently, those tools are the APIs provided and supported by nVidia and AMD/ATI for the GPU side of the fence. I know that both companies are providing support to Stanford, as best they can, but they have to cater to their customer base as well, as such, the software those APIs are based on is updating CONSTANTLY.
 
So, the way I see it, as gamers we are pushed to upgrade drivers to the newest and greatest, but as folders, Stanford would like us to stay with the tried and true.
So, if we upgrade as the nvidia or AMD software is constantly pushing us to, we start to see bugs in the process on Stanford's side of the fence. Naturally, we want to blame Stanford for instability and shoddy work.

BUT, as a scientific endeavor, they CAN NOT, NOR SHOULD NOT be forced to constantly keep pushing their methods to the 'latest/greatest/cutting edge' because that WILL introduce errors to their work that could be catastrophic.
 
So, yeah, we start to see more failures in WUs because the drivers are introducing bugs that Stanford hasn't yet accounted for. But , given time, they will.

From what I've seen, dropping back a few revisions on the drivers for Nvidia seems to have smoothed the losses out for some folks. To me, this is another sign that pulling back a little from the cutting edge is what likely needs to be done for the best results.
 
Yeah, I get frustration, especially during contests, doubly so when you can't be there to frig with your system to figure out exactly what's going on. I just have seen a LOT of heat thrown around over the years and while I also hope that there's a magical improvement in the flow from Stanford, I also know that patience is a BIG part of volunteering my services to the F@H program.
 
As I often say during rants of mine like this, this is my $.02, give or take a few points of inflation. Spend them as you will.

Way I see it, let's kick some Cancer/Alzheimer's/ALS/Chrohn's/whatever you want to target RIGHT in figurative butt, knock it down, rough it up a little bit, take it's lunch money and then tell the gang down the street that it was talking trash about their mommas.
 
d

I am using drivers 355 and haven't updated my drivers in 2 months prior to rolling back to 355. I do not do beta testing for Nvidia whql drivers like I use to. I stopped when they were constantly getting worse and stuck with the tried and true drivers so that my folding wouldn't be hampered.. To roll back from already old drivers means that Stanford my already fully stable system is now unable to complete work units that are being created with tried and true drivers that should have been stable since they were always working in the past.

If I were using updated drivers, I would agree that they are the issue with the drivers..

A response on the folding forums said he wasn't getting any lost units on 347, but of course that is win 7 or 8, so I am going to look into getting a new win 7 key strictly for a folding rig to run at all times so I can set stable drivers and leave them.

I am extremely frustrated not being able to touch my system as I would love to do error testing myself.

As of the last mode swing, I have decided to leave the system running until I get home so I can copy and paste all errors and log files when I return.

Snuck in there (the top 20) for once.
 
http://folding.extremeove...ry.php?s=&t=111065

Scarlet-Tech
The problem is that everything is set to stock now. Normally, I wouldn't stress too much about the PPD, but none of this happened until the folding challenge started. 3 days after the challenge started, I was still pushing 1.42-1.54m.. Day 4, it dropped to 1.2m. Why is it that when we start trying to crush the Stanford numbers, that suddenly an issue arrives.

I'm gonna say that you're not wrong. Both of my machines have been rock solid for months folding. I had what I expected the first 3 days of the challenge and on day 4 I also dropped about 300k ppd. The last 7 days have been a struggle just to get WU's to finish and not be returned as "bad work units". I've changed drivers and lowered video card memory speeds and still having problems. You're not wrong Scarlet, something changed on day 4

i had a p9638 fail yesterday the 10th
log file for that slot only can be found here

Gunsvnvmc
Posting anything negative at all on the Stanford Forum gets removed and then you get a nasty gram from them. So how do we get there attention and help fixing the problems on there end. They have the habit of saying everything is the fault of everyone but themselves.

the only outlet other than the forums is their subredit
https://www.reddit.com/r/foldingathome
 
Good luck with the FF Scarlet, just don't get your hopes up

Scarlet-Tech
https://foldingforum.org/...4&p=280733#p280733

So far Bruce, over at Folding forums has been replying. I am trying to relay the best information possible. It is coming down to him saying that any overclock may be the issue. I venture to guess that is going to be the only response I get honestly.

He did say this: There are reports of this sort of error with 171.64.65.56.

Problem is, he has access to information we are not able to access, so I am not sure.

Thank you un4givn85, I will share this with him, and see what he says.

We all know that there is nothing wrong with our hardware or overclocks. If anybody knows anything about system stability it is the members here. I think we all can agree that our systems are stable beyond a doubt. Just to give them the benefit of the doubt, I did lower my clocks to stock and I still had a failure on a Core x21 WU. The rate of failure didn't drop but my ppd did dramatically into the 500,000 ppd range from 750,000 ppd. I just decided I was going back to my stable OC and things have been fine without errors, warnings or bad states detected using 353,62 WHQL.

Your welcome yodap your one of the best.

Just to add to the info pile. Running 358.87 on a 960 GTX SSC 4GB with a mild OC and voltage bump. Haven't had a single WU error out on me, yet. I'll be sure to let you know if I run into one though.

My PPD plummeted the last few days because I was out of town this weekend and my client decided to act up. Then when I got home, one of my GPU's decided it didn't want to download any new WU's...just my luck.
 
Everything seems to be fixed now and she's back jugging along. 

un4givn85

bcavnaugh
Users Dropping out or just bad Projects?
 

i would say a combination of the two

Agreed

I admit I haven't been folding a whole lot the last few days, I'm still trying to diagnose the random system resets I'm having. I think it's my 1300 G2 unfortunately, I will swap it out for my spare 850 B2 later today and see if it stops.
On the bright side I haven't had any bad wu's lately.

Foldin',foldin,'foldin', keep them babies rollin', Rawhide!!!!! snap! 
I am amazed at our [team] production every time, it is on a steady increase all the time. https://sites.google.com/site/timezonechallenge/3billion
 
We are easily over the required PPD, to hammer 3 billion home. I love this folding team!
Also, thank you troy for keeping us updated and informed! :)