• 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.

Quick question

M

mrmylanman

[H]ard|Gawd
Joined
Jul 13, 2008
Messages
1,484
I recently changed OSes from Windows 7 to Windows XP x64.

In Windows 7 I was using he system tray client, and I was averaging about 6000-7000 points per day.

Nowadays, I am using the console client and Windows XP and I am getting anywhere from 2000-3000 points per day. Anyone have an idea why this could be?
 
Try the systray client for awhile and see what it gives you. Are you getting the numbers off of [H] folding or straight from the client? Sometimes the website takes a few days to average the ppd...

Also were you using any CPU clients under windows 7 that you didnt install on XP?

That is what I would do but there is a ton of smarter folders than me lurking around here...
 
Well I am using the site "kakaostats" to get my information, but that's how I got it while I was using Windows 7 too. How do you get them straight from the client? I am not too familiar with the client's settings and such, even though I have been folding for a bit now.
 
I haven't tried the monitors yet, I tried FaH Mon but it didn't work. Said something about not being able to read the log and I got annoyed by it.

Anyhow I have tried the system tray version and it's the same story. I have been paying attention to the capacitor whine my card makes, and it only makes the sound about 1/3 the time (the equivalent of if I turned the CPU usage down to about 1/3 the way). So I am not sure what the deal is. In Vista it worked just fine. Even in Windows 7.

I also verified this by looking at the viewer. It works for about 1 second, stops 2 seconds, then works 1 second. Etc.
 
It's possible that the discrepancy is caused by different units. The starting and stopping that you're describing is also characteristic of the newer core 14 workunits.
 
That's really odd. It seems strange that it'd make my PPD count go down to less than half.
 
mrmylanman said:
That's really odd. It seems strange that it'd make my PPD count go down to less than half.
Click to expand...
I tried both the systray client and the console client, and use XP but also installed Win 7 for a couple of weeks. I noticed no discernible difference in my production with any of those combinations. 2000-3000PPD is way too low for a GTX260. My 8800 GT cards are averaging around 4600PPD with the P57xx WUs and 4000PPD with the P590x WUs in XP. There must be something else going on because a halving of production is too huge a discrepancy to be solely accountable with an OS switch. You will have to check additional potential variables that could affect performance to determine a probable cause, such as video driver version. Maybe wiping out the drivers with a utility that will sweep your system to remove them, and then installing the fastest drivers will help.
 
Do you have any suggestions for drivers? I am currently using the 182.08 drivers (last I checked they are most recent). In Windows 7 I was using the most recent 7 drivers which are different so maybe that is a cause.

When I go on nVidia's site and look into CUDA they have the 181.20 drivers there, I suppose I will try that set and see if I have any luck.
 
mrmylanman said:
Do you have any suggestions for drivers? I am currently using the 182.08 drivers (last I checked they are most recent). In Windows 7 I was using the most recent 7 drivers which are different so maybe that is a cause.

When I go on nVidia's site and look into CUDA they have the 181.20 drivers there, I suppose I will try that set and see if I have any luck.
Click to expand...
The v178.24 have the best performance by reputation, but I didn't see any difference between these drivers and the v177.89 I was using prior. I suggest yo try the 178.24 driver version since it appears that from the 180s onwards the performance diminishes somewhat. Don't install the new beta drivers that have very high performance because Stanford is unsure about them.

Let us know how it goes. Best of luck. :)
 
Your point totals are too low for that card.

Have you overclocked it?
Have you set your fans to 100%?

BTW - Apollo is referring to the 185 drivers when he says don't use them. Lot of news lately because PPD goes up (sometimes a LOT) with them but 1) have been causing a lot of crashes and 2) Stanford has kind of said there might be problems with WUs returned under 185 drivers.

I'm using 182.08 in my #1 and #2 rig (both Vista) and (despite the advice) 185.2 on my folding rig (it is XP, so there was no bump from using them nor have there been any crashes, and I'm loathe to "fix" them if there is nothing wrong). My 260s are turning in about 7K PPD on box #1.
 
The video card is overclocked per the specs in my sig. The fan is not set to 100% although I don't feel it has to be, because the temps are actually pretty reasonable (around 70C.... no way it's getting throttled yet and I confirmed this through Rivatuner).

Right now I am using the 178.24 drivers (wiped the new ones off and started fresh) and it's doing the same exact thing. Kind of frustrating considering the hardware I have...
 
I'm starting to think it may have something to do with the work units I have been receiving... Right now I am folding away and I hear the constant capacitor whine (back to the console client) and it seems to be folding at the normal rate again.
 
mrmylanman said:
The video card is overclocked per the specs in my sig. The fan is not set to 100% although I don't feel it has to be, because the temps are actually pretty reasonable (around 70C.... no way it's getting throttled yet and I confirmed this through Rivatuner).
Click to expand...
Try setting the fan to 100% for a while just to see if it makes a difference.

mrmylanman said:
I'm starting to think it may have something to do with the work units I have been receiving... Right now I am folding away and I hear the constant capacitor whine (back to the console client) and it seems to be folding at the normal rate again.
Click to expand...
Hmm, well, with the slowest current crop of WUs my 8800GTs are doing 4000PPD, no less. What does your card do? Keep in mind these are much slower cards than yours is.
 
Yeah I hear ya. I haven't really gone over 3 hours now since it started getting back to normal speeds, but the fahmon (which started working now by the way) is saying 8100 PPD (I guess if it keeps at the current rate?) But my card has been consistently been getting 2400-5000 points per day for almost a week now. Before then unless I turned f@h off to play a game for a while or somethign, it was getting around 7500 PPD.
 
OK, now that's more like it! :)

If it drops below 5000PPD and no new WUs have been released, then report back here because there would still be a glitch somewhere in your setup. As it stands right now, that's good to see. :cool:
 
Yeah maybe the different drivers helped too (I am still using the 178.24 drivers) so I am hoping it holds up. I want to reach the top 1000 in hardocp before long! haha
 
Well I think I have another problem going on. And it may explain the low point counts as well. This morning when I checked the f@h log (because I saw no activity from f@hmon for a while) it said I reached the EUE limit, and it'll wait for 24 hours before giving me any more work units.

I lowered my overclock some and increased the fan speed, we shall see what happens I suppose.
 
Reached the limit again last night.

I turned all overclocking off so I guess we'll find out again in 24 hours.
 
mrmylanman said:
Reached the limit again last night.

I turned all overclocking off so I guess we'll find out again in 24 hours.
Click to expand...
Sad to see there's still a problem. You can shut down (ctrl-C) and restart the client immediately without waiting. I don't think it's OC from the looks of it, but if you fold, force the fan to operate at 100% even if it's not OC. Doing so will not only reduce the number of EUEs if the temps increase, but will lengthen the life of your card. Folding is by far the most stressful thing you can do with your GPU.
 
I gotcha. I was just running it as a service so I didn't have to mess with it, and just looked at the log file occasionally. I guess for diagnostic reasons I will check it out in a CMD window for now.

Do you know of any reasons I will produce errors? I want to get it right because there isn't much point in folding if it's all crap anyways, lol.

I will be buying another GTX260 in a couple days to SLI (mostly for gaming) at that point I know I will have to ramp the fan up for sure, but I am also curious to see what kinds of PPD I can get out if too. For that I will probably go back to Windows 7 anyway because of DirectX 10 support for games that could make use of SLI lol. Curious to see if it'll start working again purely based on that...

Very perplexing...

I checked the output just now and got a few lines that looked interesting among some others I don't know what they mean:

Code: 
[22:08:47] Assembly optimizations on if available.
[22:08:47] Entering M.D.
[22:08:53] Tpr hash work/wudata_02.tpr:  1590476833 4072469829 2293606130 110251
321 3465723787
[22:08:54] Working on Protein
[22:08:54] SEH code: 3221225477
[22:08:54] Run: exception thrown during GuardedRun
[22:08:54] Run: exception thrown in GuardedRun -- Gromacs cannot continue furthe
r.
[22:08:54] Going to send back what have done -- stepsTotalG=0
[22:08:54] Work fraction=0.0000 steps=0.
[22:08:58] logfile size=9103 infoLength=9103 edr=0 trr=23
[22:08:58] - Writing 9639 bytes of core data to disk...
[22:08:58] Done: 9127 -> 3305 (compressed to 36.2 percent)
[22:08:58]   ... Done.
[22:08:58]
[22:08:58] Folding@home Core Shutdown: UNSTABLE_MACHINE
[22:09:01] CoreStatus = 7A (122)
[22:09:01] Sending work to server
[22:09:01] Project: 5903 (Run 10, Clone 540, Gen 18)
[22:09:01] - Read packet limit of 540015616... Set to 524286976.

Essentially it keeps giving me that error "exception thrown in GuardedRun", and I am not sure what that means. Could that be what is going on?
 
Little update, I'm not sure how I fixed it but it seems to be working this time at least. I turned off the -advmethods flag and I disabled all my overclocking (video card, RAM, and CPU). I also put all 6 DIMMs in (For a while I was rolling with only 3 because it let me overclock my CPU higher and be stable).

I start Windows back up and fire up folding@home and so far it's 11% in to the first work unit and no errors (generally I have been getting errors before work units started actually working with that UNSTABLE_MACHINE line. So I guess this is a good sign for now. I have been averaging about 1% a minute so I guess that ain't too bad. I'll be keeping an eye on it for a while though to be sure.

Finished the first work unit in about an hour and a half, no errors, working on the second one now, it's taking a lot longer, but as I understand it, this work unit is worth a lot more points (if I am interpreting the "credit" field in fahmon correctly).
 
My rig is now fixed. I'm folding along at about 6500 points a day. Not too shabby. I have a second GTX260 on the way so I'll be getting a nice little boost :-D
 
You must log in or register to reply here.
Back
Top