Consolidated Team24 Rant thread

[joeechick]

My GPUs won't do more than one unit at a time

Deleted the entire folder, started from scratch, still no dice.

[15:58:19] + Processing work unit
[15:58:19] Core required: FahCore_15.exe
[15:58:19] Core found.
[15:58:19] Working on queue slot 07 [May 9 15:58:19 UTC]
[15:58:19] + Working ...
[15:58:19] - Calling '.\FahCore_15.exe -dir work/ -suffix 07 -nice 19 -priority 96 -checkpoint 15 -verbose -lifeline 2552 -version 641'

[15:58:19]
[15:58:19] *------------------------------*
[15:58:19] Folding@Home GPU Core
[15:58:19] Version 2.22 (Thu Dec 8 17:08:05 PST 2011)
[15:58:19] Build host SimbiosNvdWin7
[15:58:19] Board Type NVIDIA/CUDA
[15:58:19] Core 15
[15:58:19]
[15:58:19] Window's signal control handler registered.
[15:58:19] Preparing to commence simulation
[15:58:19] - Ensuring status. Please wait.
[15:58:28] - Looking at optimizations...
[15:58:28] - Working with standard loops on this execution.
[15:58:28] - Previous termination of core was improper.
[15:58:28] - Going to use standard loops.
[15:58:28] - Files status OK
[15:58:28] sizeof(CORE_PACKET_HDR) = 512 file=<>
[15:58:28] - Expanded 145757 -> 660994 (decompressed 453.4 percent)
[15:58:28] Called DecompressByteArray: compressed_data_size=145757 data_size=660994, decompressed_data_size=660994 diff=0
[15:58:28] - Digital signature verified
[15:58:28]
[15:58:28] Project: 8020 (Run 8, Clone 155, Gen 17)
[15:58:28]
[15:58:28] Entering M.D.
[15:58:30] Tpr hash work/wudata_07.tpr: 2072731603 3159232568 408309873 691155623 1217508881
[15:58:30] GPU device info: vendor=0 device=0 name= match=0
[15:58:30] Working on Gromacs Runs On Most of All Computer Systems
[15:58:30] Client config found, loading data.
[15:58:33] CoreStatus = 63 (99)
[15:58:33] + Error starting Folding@home core.
[15:58:33]
Folding@Home will go to sleep for 1 day as there have been 5 consecutive Cores executed which failed to complete a work unit.
[15:58:33] (To wake it up early, quit the application and restart it.)
[15:58:33] If problems persist, please visit our website at http://folding.stanford.edu for help.
[15:58:33] + Sleeping...

[18:15:46] - Autosending finished units... [May 9 18:15:46 UTC]
[18:15:46] Trying to send all finished work units
[18:15:46] + No unsent completed units remaining.
[18:15:46] - Autosend completed

[RETARD]

remotely logging in can cause an instance to EUE
check task manager to see if only one instance of fah is running
other than that ... id keep my config and delete the whole fah install
reinstall the client .... should fix the prob

[joeechick]

Originally Posted by RETARD

remotely logging in can cause an instance to EUE
check task manager to see if only one instance of fah is running
other than that ... id keep my config and delete the whole fah install
reinstall the client .... should fix the prob

Originally Posted by joeechick

Deleted the entire folder, started from scratch, still no dice.

That means the entire GPU folder, client, config, everything :P

Not remotely logging in. Will check task manager when I get home.

[The Brad]

My GPU is barfing on 7644 units with UNSTABLE_MACHINE errors.

Card is not overclocked and temp is under 70c.

Seems to be able to complete other units OK. Guess I will turn off adv-methods for now and see if it gets better?

EDIT: Fuck 30% of my total units are failed units on the GPU of this number. 2 complete 29 failed, not a single other failure in the whole system.

EDIT2: Double fuck, I do other stuff that has high CPU load so this explains why my down spikes are so down on this machine....

[mshagg]

How dare windows update restart my heavy hitter in the middle of the night

[kaz3e]

Crappy power in the west.

No power for the next 3hrs.

[The Brad]

So, with my latest electricity bill at $600 for the quarter, Folding is going to have to pause for a while.

Its a shame I've worked hard with the goal of getting into top 100 in Team24, but have seemed to plateau at around 130.

Gonna see if I can rig up a box at work where I don't have to pay for the power, but until financial things improve, this might be it.

[OhFoRkMe]

MCR-AU is going to take a very big hit shortly after my manager has requested that F@H be removed.

Since folding on our new batch of machines, we have had a number of PSU's die in a short period of time. Out of 400 machines, we have RMA'd 20 odd power supplies in those machines in the last 3 months.
It might simply be a coincidence but F@H will be removed to see if this number of dead PSU's improves.

Looking at the stats, MCR-AU was 1 Mil short of 30 Million.

Im sure I will have an opportunity to run F@H again at times with permission. How else will I be able to stress test the new 8x BL460 16 Core, 256GB Blades?

[sotech]

Originally Posted by OhFoRkMe

MCR-AU is going to take a very big hit shortly after my manager has requested that F@H be removed.

Since folding on our new batch of machines, we have had a number of PSU's die in a short period of time. Out of 400 machines, we have RMA'd 20 odd power supplies in those machines in the last 3 months.
It might simply be a coincidence but F@H will be removed to see if this number of dead PSU's improves.

Looking at the stats, MCR-AU was 1 Mil short of 30 Million.

Im sure I will have an opportunity to run F@H again at times with permission. How else will I be able to stress test the new 8x BL460 16 Core, 256GB Blades?

Yikes, those would have some serious PPD...

I've just managed to start folding again after a 4-5 month sabbatical - not much output at this stage but it'll hopefully improve over the next few weeks if I can set it up on another couple of computers here :/ Power bill ended up getting too much at home and had to cut back for a while.

[dazzawul]

Well, last power bill was 2 grand for the quarter... luckily we're about to get a 5kW solar array installed, but fuck me, last couple of upgrades really upped the power bill hey :|

[sotech]

Originally Posted by dazzawul

Well, last power bill was 2 grand for the quarter... luckily we're about to get a 5kW solar array installed, but fuck me, last couple of upgrades really upped the power bill hey :|

Well, damn. I stopped for while once it hit $700/quarter to make some changes to fold more power efficiently... I think my wife would sell the folding boxes while I wasn't looking if we got a bill that big

[The Brad]

I'm so hoping to be able to resume folding in a limited way once my first post-solar power bill comes in, but in the meantime I'm having a horrid couple of months, nearly $10k worth of exceptional failures/repairs at the house, and then some fucker vandalises my car, cuts up the leather seats and dash with the broken glass... Wish I could sell a kidney!

[trevor68]

Well I built a new box last month, 3570K on a Asrock Extreme4. Its been folding like a train, doing around 25k a day.

On wednesday I added another 8 gig of matching G.Skill 1866 RAM, for a total of 16gig.

It hasn't been able to complete a single unit since.

[Lucus Maximus]

Originally Posted by trevor68

Well I built a new box last month, 3570K on a Asrock Extreme4. Its been folding like a train, doing around 25k a day.

On wednesday I added another 8 gig of matching G.Skill 1866 RAM, for a total of 16gig.

It hasn't been able to complete a single unit since.

The RAM Timings might not be stable with the extra sticks in there maybe. Or might need to up some volts somewhere. Otherwise back down some timings and see how it goes.

-Lucus

[trevor68]

I tried upping the volts quite a bit with no success.

The ram is G.Skill rated at 1866 @ 1.5 volts on their webby.

I will run some memtest tomorrow to see if its a dodgy stick perhaps.

If all else fails I'll just go back to 8 gig, that seemed ample for gaming to me anyhow.