• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

WTF?!?! 5801 WU

M

MrBooMY

Gawd
2FA
Joined
Sep 27, 2004
Messages
991
I have one computer that started getting 5801 WUs. They keep saying my card is unstable, so I clocked it down to the STOCK clocks. They are STILL saying my card is unstable. What gives?!?!??! This is BS. It wasn't having ANY trouble with the other blocks. And for some reason it just KEEPS assigning them to me. FU Stanford FU!!!!
 
im still doing 5506 WU's.. been doing 5506 and 5516 WU's for the last couple days..
 
My other computers were, but now my main computer got a 5801 then a THANK GOD 5800.
 
Same thing here as well. Half my GPU clients already reached the EUE limit and pausing for 24hrs. :(

 
MrBooMY said:
I have one computer that started getting 5801 WUs. They keep saying my card is unstable, so I clocked it down to the STOCK clocks. They are STILL saying my card is unstable. What gives?!?!??! This is BS. It wasn't having ANY trouble with the other blocks. And for some reason it just KEEPS assigning them to me. FU Stanford FU!!!!
Click to expand...

Wow. Must you be so dramatic and bitter about everything? You've been around a while now. Servers down? Bad WUs released? Been there, done that. This is SOP for Stanford. Lighten up a little bit. Swearing at Stanford and complaining on the boards isn't helping anyone.:rolleyes: If you dont like the way things are going with the project then do something about it, dont just complain about it. The WCG client has been 100% stable for me thus far. ;)

For what its worth. I've had the same resultsfor the 5800s on my 8800GS that you have experienced with the 5801s. Most of the 5800s result in a UNSTABLE_MACHINE error on my OC that has been stable for all other Nvidia WUs. Dont think I've had any 5801s yet.
 
Yeah, what gives :confused: After the aggravating BS of waiting for a WU I get a 1801 WU on my eVGA 9600GSO (finally screwed up :(, never, ever had a problem with it :D) I get the "unstable machine" shite in the console so I think "maybe the shaders are OC'ed too high" so I just set the shaders back to 1375 (which the nVIDIA web page says is stock). Low and "f*cking behold" I get another "unstable machine".

I can't speak for anyone but myself, but if this is an example of the Stanford administration mind set (ie: no work, bogus Stanford's "jerk off" Web page, unstable WU's and unstable WinSMP clients, etc) then I'll just convert my boxen over to WCG and just say f*uck Stanford, the F@H project and the worm they rode in on (errr....., horse, mule, llama or something) I ain't never, ever had a problem with the WCG clients. Hell, I might check out the GPU deal for the Grid deal (or something like that) :p

I realize this doesn't bother Stanford in the least (hell, they gots plenty of suckers), but I believe in that old sayin' "you can fool some of the people some of the time, but you can't fool all of the people all of the time" (or something like that)

After writing this post I'm gonna just shut this boxen down for a while and try later :mad:

Folding and WCGing for a CURE

 
metallicafan said:
Wow. Must you be so dramatic and bitter about everything? You've been around a while now. Servers down? Bad WUs released? Been there, done that. This is SOP for Stanford. Lighten up a little bit. Swearing at Stanford and complaining on the boards isn't helping anyone.:rolleyes: If you dont like the way things are going with the project then do something about it, dont just complain about it. The WCG client has been 100% stable for me thus far. ;)

For what its worth. I've had the same resultsfor the 5800s on my 8800GS that you have experienced with the 5801s. Most of the 5800s result in a UNSTABLE_MACHINE error on my OC that has been stable for all other Nvidia WUs. Dont think I've had any 5801s yet.
Click to expand...

Correct, a better way to express frustration is to write a email to Vijay. I just did that 5 mins ago in a polite tone asking him to stop everything and get the whole group sit together to assess the situation. They are chasing too many rabbits lately and let the stove in a critical situation.

 
I'm having the same problem as others, both GPU clients giving the message of death. Just shut them down until they have it fixed. The last 4 days have been the shits.....

 
Pooch666 said:
I'm having the same problem as others, both GPU clients giving the message of death. Just shut them down until they have it fixed. The last 4 days have been the shits.....
Click to expand...
I wouldn't shut them down just yet. Most of my GPU clients stopped because they exceeded the EUE limit, but I just restarted them and they received P5506 WUs instead of the accursed P5801. They're working for the time being.
 
I've shut down my GPU client. I got tired of running 90 MPH and getting nowhere. My GPU client has spent 50% of the last several days inactive waiting for a WU. Then it gets a couple of WUs that cause total instability and EUEs (I honestly can't count the number today due to this WU). I give up and will not turn it on until this sh!t is stable.

 
APOLLO said:
I wouldn't shut them down just yet. Most of my GPU clients stopped because they exceeded the EUE limit, but I just restarted them and they received P5506 WUs instead of the accursed P5801. They're working for the time being.
Click to expand...

I restarted 3 times in the hopes that it would clear the "Exceeded EUE limit...". Nada, kept coming up. I give up
 
I finally got a 5015, hopefully I can get SOMETHING done today. :D
 
capreppy said:
I've shut down my GPU client. I got tired of running 90 MPH and getting nowhere. My GPU client has spent 50% of the last several days inactive waiting for a WU. Then it gets a couple of WUs that cause total instability and EUEs (I honestly can't count the number today due to this WU). I give up and will not turn it on until this sh!t is stable.

I restarted 3 times in the hopes that it would clear the "Exceeded EUE limit...". Nada, kept coming up. I give up
Click to expand...
I fully understand and you're not the only one feeling this way. I haven't seen it this bad in very long time, perhaps years. Part of the problem is the unheralded popularity of the nVidia client. A cheap and easy way to get big production and everyone jumped on the bandwagon. Couple that with the recent bad GPU core release and it's a recipe for disaster. I'm dealing with this issue plus problems with the console and SMP clients. The only thing I haven't dealt with are PS3 client-related problems only because I don't have one.
 
APOLLO said:
I fully understand and you're not the only one feeling this way. I haven't seen it this bad in very long time, perhaps years. Part of the problem is the unheralded popularity of the nVidia client. A cheap and easy way to get big production and everyone jumped on the bandwagon. Couple that with the recent bad GPU core release and it's a recipe for disaster. I'm dealing with this issue plus problems with the console and SMP clients. The only thing I haven't dealt with are PS3 client-related problems only because I don't have one.
Click to expand...

I wouldn't say years. We had that PS3 situation a a few months back where we could barely download or upload WU's for a couple weeks. I might have gotten a unit downloaded a day, and these were 3 hour 125 point units, and half of them would expire in the queue with their 2 day deadlines.

Good times, good times. :D


2.png
 
alan2308 said:
I wouldn't say years. We had that PS3 situation a a few months back where we could barely download or upload WU's for a couple weeks. I might have gotten a unit downloaded a day, and these were 3 hour 125 point units, and half of them would expire in the queue with their 2 day deadlines.
Click to expand...
Yeah, I remember all the bitching about that. Some people were contemplating selling their PS3s because the problem took so long to resolve. Bad as it was, this juncture we find ourselves in is unique in the sense of the confluence of varied simultaneous events that have conspired to cripple the DC project on many fronts. I've had GPU, SMP and console client server problems in the last 24 hours. The GPU server problem can be termed chronic, and the GPU core EUE issue standing for well over a week as of this writing. In that sense this particular situation is more far reaching in its scope than other Stanford debacles of the past.

 
APOLLO said:
Yeah, I remember all the bitching about that. Some people were contemplating selling their PS3s because the problem took so long to resolve. Bad as it was, this juncture we find ourselves in is unique in the sense of the confluence of varied simultaneous events that have conspired to cripple the DC project on many fronts. I've had GPU, SMP and console client server problems in the last 24 hours. The GPU server problem can be termed chronic, and the GPU core EUE issue standing for well over a week as of this writing. In that sense this particular situation is more far reaching in its scope than other Stanford debacles of the past.

Click to expand...

Yeah, it looks like they are sucked into a perfect storm :(

The bugger is that even if the servers is under control of the PG, it's on the network rooms managed by a independent Stanford IT team and there is no communication between them. Imagine them yanking the plug and moving them in a inopportun timing. Today's outage is one of them. Vijay told me it take lots of money to have their own room and their own servers with a sysadmin so it's not a easy solution (he also told me that they basically need double the current servers count to absorb the load).

I told them to put the brakes on any future new project and to work to get everything stable. When this happen, they can restart slowly and surely to get things working fine.

 
Xilikon said:
Yeah, it looks like they are sucked into a perfect storm :(

The bugger is that even if the servers is under control of the PG, it's on the network rooms managed by a independent Stanford IT team and there is no communication between them. Imagine them yanking the plug and moving them in a inopportun timing. Today's outage is one of them. Vijay told me it take lots of money to have their own room and their own servers with a sysadmin so it's not a easy solution (he also told me that they basically need double the current servers count to absorb the load).
Click to expand...
Hmm, the situation is far worse than I imagined then. That's very unfortunate. It sounds like Vijay is admitting that Stanford is way over its head. Aside from a dedicated server room and a sysadmin, what else can be accomplished to alleviate the problems? What about additional hardware, through donations for example? This isn't the same as the complete solution he suggested for a wish list, but will that help in any way?

I told them to put the brakes on any future new project and to work to get everything stable. When this happen, they can restart slowly and surely to get things working fine.
Click to expand...
Excellent idea. Hopefully, they'll abide by your advice. ;)
 
Remember that at the same period last year, we just celebrated breaking the 1 Petaflop barrier. We are now close to 5 petaflops so it increased 5x without the backend following the increase, hence the increasing issues.

Without the money and hardware to improve the backend, the only way is to have things working stable (lots of EUE often mean more traffic). I'm discussing with Vijay to find ways to improve the QA process since it seems to have failed a lot at catching problems before it get in the wild and I'm trying hard to give them good ideas.

I'm sure they already considered donations but remember that we also need infrastructures to handle the traffic and Stanford campus seems unable to provide a reliable infrastructure :(

 
My clients are sporatically picking up work. Nothing to get excited about. Since the overall cost of idle clients to me is only a couple of bucks for electricity, it's no big deal. Everybody who's getting upset, take a deep breath, count to 10 and move on. If the points race means that much, just figure that most likely nobody is gaining much if any ground on you.

Now if we were competing for cash prizes....



 
Yeah... Last 2 days the servers were down, now I'm at my 24 hour EUE limit. Fucking great.
 
I just read through the FCF thread regarding 5801. People are just shutting down their GPU clients until such a time as this mess is sorted out and the servers recover.

I had planned out 2 boxen with 2 9800GX2's each. I was planning to purchase and get these up next week. I'm sorry, but I am now re-evaluating how much I want to put into folding.

Yes it is about the science, but let's face it, I'm a points whore. I AM about the points. I don't want to purchase equipment if there are stability issues. The boxen aren't cheap and electricity even less so.

 
all you have to do is close and restart it.. and it will download a different WU..
 
7 GPU all down when I came home from work, one was up :mad:

 
5801 here. Like everybody else,,, it won't run. NAN's right out.

So my gtx260 and 8800gt are out of the game. When my 8800gts gets one and fails I will shut that machine down also. Im not even going to bother with putting smp or other clients on my machines.

While I am not a points whore. I am starting to see the antics coming out of the left coast as pure lunacy. They can't keep servers up, the clients and cores are turning into a mess and the work units are a waste of bandwidth and electricity.

I will keep my quad running smp as long as it will go on the old reissued core. If they do not get things straight. I will stop folding.

Sorry team. Im just about done with the idiots and there stupid games.
 
I PMed Vijay and they will pull all of them off asap. I just told them to do one thing at a time and right now, we should concentrate to get a new nvidia core out which work better than currently.

 
I have no issues with the new core. It runs great on my boxen. Ive even ran the 5800 unit with out issue. But they don't seem to see the 30% or so that had massive failure or better off the bat? Even a 3rd grader could see shit was going way wrong real quick. My little grandson saw it for heavens sake.

I just wonder how many baseball bats it takes to beat common sense into over educated smart folks. Do they not see the disaster they have caused? Are they blind?

Or do they think folks are stupid.
 
5800s work for me as well, the 5801s are the problem. This is the first work unit that I've had problems with. It crashed ALL of my GPUs today, including the ones at my friend's house. I just restarted 2 of my 4 at my house and luckily got 5800s on both. My points are suffering horribly today. Damn you Stanford!! Did they even TEST the damn units before they put them out. If they did, this wouldn't be happening.
 
MrBooMY said:
5800s work for me as well, the 5801s are the problem. This is the first work unit that I've had problems with. It crashed ALL of my GPUs today, including the ones at my friend's house. I just restarted 2 of my 4 at my house and luckily got 5800s on both. My points are suffering horribly today. Damn you Stanford!! Did they even TEST the damn units before they put them out. If they did, this wouldn't be happening.
Click to expand...

They did the test. However, when they tested it, they used a older core and everything ran great. However, they didn't know it's behaving horribly with core 1.15 and newer so they pulled them off for now.

 
I still don't see why it took hours for them to pull this WU. Thats completely unacceptable by any means. It was still w/in normal working hours in Cal when the crap in the fan.

Remember that at the same period last year, we just celebrated breaking the 1 Petaflop barrier. We are now close to 5 petaflops so it increased 5x without the backend following the increase, hence the increasing issues.
Click to expand...

Because folding power has increase by 5x doesnt mean the bandiwidth need has increased by 5x. Most of the increase is from GPU2 and those WUs are very minimal in size to download unlike SMP. My log shows 46k to download, 1 sec at 44kB/s. SMP WUs take several minutes to upload and I can do 250kB/s Up. Releasing 15 point WUs causes resource problems...
 
That is stupidity right there. Why test it with something old?!?! That makes no sense to me. That was the true cause of this problem.
 
MrBooMY said:
That is stupidity right there. Why test it with something old?!?! That makes no sense to me. That was the true cause of this problem.
Click to expand...
I fully concur here especially since core v1.15 was a forced DL and newer WUs won't run without it.
 
APOLLO said:
Hmm, the situation is far worse than I imagined then. That's very unfortunate. It sounds like Vijay is admitting that Stanford is way over its head. Aside from a dedicated server room and a sysadmin, what else can be accomplished to alleviate the problems? What about additional hardware, through donations for example? This isn't the same as the complete solution he suggested for a wish list, but will that help in any way?

Excellent idea. Hopefully, they'll abide by your advice. ;)
Click to expand...

Ah grasshopper you are learning. After years of frustration, loud verbal explosions, high blood pressure, extreme angst, a bit of adgeta and ailments too numerous to mention I finally said, very clearly I might add; FUCK IT.

Then I took, or tried to take a kinder gentler approach which would be my true nature anyway and contacted VJ directly. In addition I did my best to work with 7im over at the Community Forum. All during my last go round Xil and I were talking about things we could to do improve things.

Again I reached the FUCK IT level, however credit must be given here to our own Xilikon for carrying on despite the mental and possible physiological damage he might sustain. Face it, he is French(ish) and I frankly didn’t think anyone would notice the difference.;)

Through all this I still continue to fold because I believe it is a genuine noble cause and I am one noble guy.

Yes, for those that noticed, my numbers have been way off this month. About once a year or so I do take a mental health break. This month I took out all my frustrations out in Soap Box here on the forum and poor “Barry” has I hope suffered at my hands. ;)

My point is, don’t let this crap get to you. Clearly the cause is worth it but clearly VJ has no clue how to run it. I should start producing numbers again by the first of the month or there abouts.

Oops, I have “Trading Places” on the big TV and Jamie Lee just took her top off…..

Fold on grasshopper and heed relic’s comment in my sig re VJ;)




 
You must log in or register to reply here.
Back
Top