New Members Welcome & Advice Thread

Xroger,
that log is a single core client log. you probably need smp client on that phenom... have to use the smp flag.... delete your clients and setup smp...
the guides are linked next to the dl at fah http://folding.stanford.edu/English/DownloadWinOther (smp is at the bottom).

Nemon,
all you need to do to join xs folding is start folding and use team# 36362 during your config when you run the client for the first time.
Vista 64 bit, going from the hardware in your sig...
run the nvidia gpu systray client http://folding.stanford.edu/English/DownloadWinOther
and since you're on a intel quad, run dual smp with AC.. here's a quickie walkthru http://www.xtremesystems.org/forums/...d.php?t=199283

xroger, you can use that walkthru for smp too, but I'm not sure if AC or dual smp applies to a phenom... but you can still use the walkthru for a qucker way than fah's long guide method. just skip AC and only do one client.


welcome to the team guys... :)

Quote:

---

Originally Posted by MikeB12

Xroger,
that log is a single core client log. you probably need smp client on that phenom... have to use the smp flag.... delete your clients and setup smp...
the guides are linked next to the dl at fah http://folding.stanford.edu/English/DownloadWinOther (smp is at the bottom).

xroger, you can use that walkthru for smp too, but I'm not sure if AC or dual smp applies to a phenom... but you can still use the walkthru for a qucker way than fah's long guide method. just skip AC and only do one client.


welcome to the team guys... :)

---

well i guess my system does like it , im sry guys i cant do it cause everytime i look to see how far i have gotten it starts over , talking about being depressed ,i cant help and that upsetting

Most of us use Fahmon to see how the folding is doing http://fahmon.net/ Maybe if you do too then you can keep going although I've not had your problem so can't really help as to why you have this issue

I would think that the Phenom X4 would be fantastic for dual-SMP...not sure about the X3 though. You can use the WinAFC program to load-balance the SMP threads...

Quote:

---

Originally Posted by MikeB12

Xroger,
that log is a single core client log. you probably need smp client on that phenom... have to use the smp flag.... delete your clients and setup smp...
the guides are linked next to the dl at fah http://folding.stanford.edu/English/DownloadWinOther (smp is at the bottom).

Nemon,
all you need to do to join xs folding is start folding and use team# 36362 during your config when you run the client for the first time.
Vista 64 bit, going from the hardware in your sig...
run the nvidia gpu systray client http://folding.stanford.edu/English/DownloadWinOther
and since you're on a intel quad, run dual smp with AC.. here's a quickie walkthru http://www.xtremesystems.org/forums/...d.php?t=199283

xroger, you can use that walkthru for smp too, but I'm not sure if AC or dual smp applies to a phenom... but you can still use the walkthru for a qucker way than fah's long guide method. just skip AC and only do one client.


welcome to the team guys... :)

---

Thanks, that tutorial worked great!
Question: Can my system run both smp's and the GPU client?

Which is better? 2xsmp or 1x GPU?

thanks

Quote:

---

Originally Posted by Nemon

Thanks, that tutorial worked great!
Question: Can my system run both smp's and the GPU client?

Which is better? 2xsmp or 1x GPU?

thanks

---

You running Vista or XP?
With Vista and a quad you can run 2smp and the GPU no problem. With that i7 I dont know it may work just as well in XP.

yeah Nemon, on vista64 and your intel quad, keep the dual smp clients running...
and install the gpu systray client for your 260 (much easier setup than smp).
http://www.stanford.edu/group/pandeg...ystray-623.msi

all you need to do is put your username and team # 36362 in the client, which can be accessed by right clicking the systray icon it installs and configure. it's gui... if you want the shortcut on your desktop instead of the startup folder. just move the gpu icon from your windows start menu-startup folder to the desktop. that's how I do mine, so windows wont try and start the gpu client every boot. I can choose when to run it.

then to monitor them install a 3rd party app fahmon. http://fahmon.net/
and right click the white area to add your clients to monitor.
the smp clients just get pointed to the install dir's.
the gpu client is in your profile, my path is at the bottom of the window here.
this is one of my Q6600's and a 8800gt on 2 smp's and a gpu client...
pretty much the same as you except, you have better hardware.. (but the config is the same - intel quad and one nvidia gpu). btw: don't use the viewer in the gpu client, just use fahmon...

:yepp:
:welcome:

you know i just thought of something about folding here i am ready to give up cause im having this problem with folding , but wtf it like i will be giving up on the people that we are helping and i cant do that ,i might not know how to install this but damn you dont see the people that have this illness giving up ,im ashamed of myself

anyways sry for that small rant but anywho ,i have read alot but i still cant get this going only in single core i want to atleast run two cores and i want to see everything so when i goto the ttaskbar i can do what i want to do
i cant understand all the technical talk, would anyone be able to walk me though this

Also again i will be folding with the with the system thats in my sig

Thanks Roger

OK, here let's start from scratch :)

Remove everything FAH related first.

1. Download FAH SMP HERE

2. Install it. Preferably outside Program Files, so somewhere like C:\FAH SMP

3. You'll need a password on your user account.
If you are running Vista, you'll need to turn off UAC and reboot for the change to take effect. I've been told that you don't have to, just run as admin, but I know disabling UAC works and UAC annoys me anyway :p:

4. Go into where you installed it, then run install.bat. It should have your username already, just hit enter, then type in your password when prompted. I believe it won't display anything as you type, so type slowly and carefully so as to not mess up. I think it asks you twice. After you do that, it should say "If you see this, MPI is working" or something like that. If so, great! next step. If not, try running it again and retype your password.

5. In XP, go to Run and type in (including quotes) "C:\FAH SMP\folding@home-win32-x86.exe" -configonly (assuming C:\FAH SMP is where you installed it to). Put in your username, your team number 36362, your passkey is probably blank so just hit enter, no to sending/fetching work, no to proxy, big for acceptable work unit size, yes to change advanced options, no to launch automatically as a service (this is how I run, so there is a button on the taskbar), core priority idle, 100 CPU requested, no to disable optimized code, no to pause if on battery power, whatever you want for checkpoints (I use 5 min), whatever memory to indicate (I typically use full amount), yes to set -advmethods flag always, no to ignore deadline info, 1 for machine ID, no for disabling CPU affinity lock, nothing for additional client parameters, nothing for IP address core binding.

6. Create a shortcut, say on your desktop, with this target (including quotes): "C:\FAH SMP\Folding@home-Win32-x86.exe" -smp -forceasm -verbosity 9
-smp tells it to run in SMP mode. -forceasm I've found helps with keeping it from going bonkers when restarting the client. -verbosity 9 makes it log everything that happens, makes troubleshooting easier.

That is how I set it up, and it works. If you need any more help or clarification on these steps, let me know! :)

ok i have tryed this like 5times and this is what i get

Stopping MPICH2 Process Manager, Argonne National Lab.
MPICH2 Process Manager, Argonne National Lab stopped.
MPICH2 Process Manager, Argonne National Lab removed.
MPICH2 Process Manager, Argonne National Lab installed.
account (domain\user) [RAID0VIDEOOC1\Roger]: XRogerX
password:
confirm password:
Password encrypted into the Registry.
Credentials for XRogerX rejected connecting to raid0VideoOC1
Aborting: Unable to connect to raid0VideoOC1
Press any key to continue . . .

For the account, put in what you see in those brackets. RAID0VIDEOOC1\Roger
That is your account ID on your PC, not your F@H username :)

When you run the config part in step 5, that is where you put in your F@H user ID :)

ok this is what i got now


Note: Please read the license agreement (Folding@home-Win32-x86.exe -license). F
urther
use of this software requires that you have read and accepted this agreement.



--- Opening Log file [December 17 19:24:46 UTC]


# Windows CPU Console Edition #################################################
################################################## #############################

Folding@Home Client Version 6.22 SMP Beta2

http://folding.stanford.edu

################################################## #############################
################################################## #############################

Launch directory: C:\FAH SMP
Executable: C:\FAH SMP\Folding@home-Win32-x86.exe


[19:24:46] - Ask before connecting: No
[19:24:46] - User name: XRogerX (Team 36362)
[19:24:46] - User ID: 51E19A292C3F3E86
[19:24:46] - Machine ID: 1
[19:24:46]
[19:24:47] Loaded queue successfully.
[19:24:47]
[19:24:47] + Processing work unit
[19:24:47] Core required: FahCore_78.exe
[19:24:47] Core found.
[19:24:47] Working on queue slot 01 [December 17 19:24:47 UTC]
[19:24:47] + Working ...
[19:24:47]
[19:24:47] *------------------------------*
[19:24:47] Folding@Home Gromacs Core
[19:24:47] Version 1.90 (March 8, 2006)
[19:24:47]
[19:24:47] Preparing to commence simulation
[19:24:47] - Looking at optimizations...
[19:24:47] - Files status OK
[19:24:47] - Expanded 238214 -> 1167721 (decompressed 490.1 percent)
[19:24:47]
[19:24:47] Project: 4440 (Run 592, Clone 0, Gen 1)
[19:24:47]
[19:24:47] Assembly optimizations on if available.
[19:24:47] Entering M.D.
[19:25:07] (Starting from checkpoint)
[19:25:07] Protein: p4440_Seq42_Amber03
[19:25:07]
[19:25:07] Writing local files
[19:26:25] Extra SSE boost OK.
[19:26:25] Writing local files
[19:26:25] Completed 0 out of 1500000 steps (0%)


ok now i dont understand the desktop short cut , i think im doing the wrong this but what do i know

Well without setting up the desktop shortcut right you won't get SMP running, you're back on uniprocessor mode.

I attached a picture of what the shortcut target line should look like. You can't see the whole thing but the main part of interest is visible. Shut it down. Change your desktop shortcut to look like that (just add the stuff on after the quote) But leave your Start in alone. Mine is located in a different spot. Then go into that folder and delete the work folder and the queue.dat file. Start it up again.

Cant get both SMP's to run. FAH SMP Affinity Changer 64bit is installed and running.
Task manager says both smp's are running, FAH MON shows ony one

Aslo GPU will not show up on FAH MON

Help on fixing this.

Thanks

http://i28.photobucket.com/albums/c239/vkvenom/fah.jpg

ok, the ppd on that smp looks like a single core unit.
stop the clients, delete the work folder and queue.dat file from your smp install directories to get rid of the single core untis.
you need to follow that walkthru to the letter, including adding the flags in the shortcut. http://www.xtremesystems.org/forums/...d.php?t=199283

make sure you have typed the smp and local flags with a space between in the desktop shortcuts,

Code:

---

Win32-x86.exe -smp -local

---

this tells the client to run in smp mode, instead of single core mode.. then when you launch the smp clients you should have 8 instances of fahcore_a1 in taskmgr. 4 per client.

to read your gpu in fahmon. you need to point it to the path example in my ss below. the path at the bottom shows you where. you will nee to show hidden files and folders to see it - from folder options in ctrl panel

http://www.xtremesystems.org/forums/...1&d=1229515355

Quote:

---

Originally Posted by SparkyJJO

Well without setting up the desktop shortcut right you won't get SMP running, you're back on uniprocessor mode.

I attached a picture of what the shortcut target line should look like. You can't see the whole thing but the main part of interest is visible. Shut it down. Change your desktop shortcut to look like that (just add the stuff on after the quote) But leave your Start in alone. Mine is located in a different spot. Then go into that folder and delete the work folder and the queue.dat file. Start it up again.

---

wow i made the shortcut but it showing that im still running single core

Note: Please read the license agreement (Folding@home-Win32-x86 -smp -f -license
). Further
use of this software requires that you have read and accepted this agreement.




--- Opening Log file [December 17 22:40:33 UTC]


# Windows CPU Console Edition #################################################
################################################## #############################

Folding@Home Client Version 6.22 SMP Beta2

http://folding.stanford.edu

################################################## #############################
################################################## #############################

Launch directory: C:\FAH SMP
Executable: C:\FAH SMP\Folding@home-Win32-x86 -smp -forceasm -verbosity 9.exe


[22:40:33] - Ask before connecting: No
[22:40:33] - User name: XRogerX (Team 36362)
[22:40:33] - User ID: 51E19A292C3F3E86
[22:40:33] - Machine ID: 1
[22:40:33]
[22:40:33] Loaded queue successfully.
[22:40:33] - Preparing to get new work unit...
[22:40:33] + Attempting to get work packet
[22:40:33] - Connecting to assignment server
[22:40:34] - Successful: assigned to (171.67.108.13).
[22:40:34] + News From Folding@Home: Welcome to Folding@Home
[22:40:34] Loaded queue successfully.
[22:40:45] + Closed connections
[22:40:45]
[22:40:45] + Processing work unit
[22:40:45] Core required: FahCore_78.exe
[22:40:45] Core found.
[22:40:45] Working on queue slot 01 [December 17 22:40:45 UTC]
[22:40:45] + Working ...
[22:40:45]
[22:40:45] *------------------------------*
[22:40:45] Folding@Home Gromacs Core
[22:40:45] Version 1.90 (March 8, 2006)
[22:40:45]
[22:40:45] Preparing to commence simulation
[22:40:45] - Looking at optimizations...
[22:40:45] - Created dyn
[22:40:45] - Files status OK
[22:40:45] - Expanded 238048 -> 1168241 (decompressed 490.7 percent)
[22:40:45] - Starting from initial work packet
[22:40:45]
[22:40:45] Project: 4438 (Run 602, Clone 0, Gen 1)
[22:40:45]
[22:40:45] Assembly optimizations on if available.
[22:40:45] Entering M.D.
[22:40:52] Protein: p4438_Seq41_Amber03
[22:40:52]
[22:40:52] Writing local files
[22:42:10] Extra SSE boost OK.
[22:42:11] Writing local files
[22:42:11] Completed 0 out of 1500000 steps (0%)




Im also getting this to



4 cores detected
Credentials for RAID0VIDEOOC1\Roger rejected connecting to raid0VideoOC1
Aborting: Unable to connect to raid0VideoOC1
[22:27:41] Error running mpiexec. Please check your configuration and re-run
stall.bat if necessary

stop the client and delete the enire work folder and queue.dat file from your install dir. until you do this, it will try to finish that single core unit you already started.

when you launched install.bat.. did you get "mpi is working" in the message.
the user you use for install.bat needs to be a windows admin. not your folding user.

Ok i got it working on all 4cores here is the long of it

Note: Please read the license agreement (Folding@home-Win32-x86.exe -license). F
urther
use of this software requires that you have read and accepted this agreement.

4 cores detected
If you see this twice, MPI is working
If you see this twice, MPI is working


--- Opening Log file [December 17 23:16:36 UTC]


# Windows SMP Console Edition #################################################
################################################## #############################

Folding@Home Client Version 6.22 SMP Beta2

http://folding.stanford.edu

################################################## #############################
################################################## #############################

Launch directory: C:\FAH SMP
Executable: C:\FAH SMP\Folding@home-Win32-x86.exe
Arguments: -smp -forceasm -verbosity 9

Warning:
By using the -forceasm flag, you are overriding
safeguards in the program. If you did not intend to
do this, please restart the program without -forceasm.
If work units are not completing fully (and particularly
if your machine is overclocked), then please discontinue
use of the flag.

[23:16:36] - Ask before connecting: No
[23:16:36] - User name: XRogerX (Team 36362)
[23:16:36] - User ID: 51E19A292C3F3E86
[23:16:36] - Machine ID: 1
[23:16:36]
[23:16:36] Work directory not found. Creating...
[23:16:36] Could not open work queue, generating new queue...
[23:16:36] - Preparing to get new work unit...
[23:16:36] - Autosending finished units... [December 17 23:16:36 UTC]
[23:16:36] + Attempting to get work packet
[23:16:36] Trying to send all finished work units
[23:16:36] - Will indicate memory of 3036 MB
[23:16:36] + No unsent completed units remaining.
[23:16:36] - Detect CPU.[23:16:36] - Autosend completed
Vendor: AuthenticAMD, Family: 15, Model: 2, Stepping: 3
[23:16:36] - Connecting to assignment server
[23:16:36] Connecting to http://assign.stanford.edu:8080/
[23:16:36] Posted data.
[23:16:36] Initial: 40AB; - Successful: assigned to (171.64.65.64).
[23:16:36] + News From Folding@Home: Welcome to Folding@Home
[23:16:36] Loaded queue successfully.
[23:16:36] Connecting to http://171.64.65.64:8080/
[23:16:40] Posted data.
[23:16:40] Initial: 0000; - Receiving payload (expected size: 2428300)
[23:16:42] - Downloaded at ~1185 kB/s
[23:16:42] - Averaged speed for that direction ~1185 kB/s
[23:16:42] + Received work.
[23:16:42] + Closed connections
[23:16:42]
[23:16:42] + Processing work unit
[23:16:42] Work type a1 not eligible for variable processors
[23:16:42] Core required: FahCore_a1.exe
[23:16:42] Core not found.
[23:16:42] - Core is not present or corrupted.
[23:16:42] - Attempting to download new core...
[23:16:42] + Downloading new core: FahCore_a1.exe
[23:16:42] Downloading core (/~pande/Win32/x86/Core_a1.fah from www.stanford.edu
)
[23:16:43] Initial: AFDE; + 10240 bytes downloaded
[23:16:43] Initial: AD21; + 20480 bytes downloaded
[23:16:43] Initial: CC38; + 30720 bytes downloaded
[23:16:43] Initial: 8501; + 40960 bytes downloaded
[23:16:44] Initial: F56A; + 51200 bytes downloaded
[23:16:44] Initial: ABAE; + 61440 bytes downloaded
[23:16:44] Initial: B6B0; + 71680 bytes downloaded
[23:16:44] Initial: 783A; + 81920 bytes downloaded
[23:16:44] Initial: B2A6; + 92160 bytes downloaded
[23:16:44] Initial: 1409; + 102400 bytes downloaded
[23:16:44] Initial: BBF0; + 112640 bytes downloaded
[23:16:44] Initial: 1861; + 122880 bytes downloaded
[23:16:44] Initial: 5950; + 133120 bytes downloaded
[23:16:44] Initial: 1081; + 143360 bytes downloaded
[23:16:44] Initial: 26BC; + 153600 bytes downloaded
[23:16:44] Initial: FE4A; + 163840 bytes downloaded
[23:16:44] Initial: C1C3; + 174080 bytes downloaded
[23:16:44] Initial: 9B49; + 184320 bytes downloaded
[23:16:44] Initial: 9EE5; + 194560 bytes downloaded
[23:16:44] Initial: D79D; + 204800 bytes downloaded
[23:16:44] Initial: 7801; + 215040 bytes downloaded
[23:16:44] Initial: 8B51; + 225280 bytes downloaded
[23:16:44] Initial: E26E; + 235520 bytes downloaded
[23:16:44] Initial: EDB0; + 245760 bytes downloaded
[23:16:44] Initial: 0919; + 256000 bytes downloaded
[23:16:44] Initial: CDDE; + 266240 bytes downloaded
[23:16:44] Initial: 7A7E; + 276480 bytes downloaded
[23:16:44] Initial: 034E; + 286720 bytes downloaded
[23:16:44] Initial: 88D0; + 296960 bytes downloaded
[23:16:44] Initial: D66D; + 307200 bytes downloaded
[23:16:44] Initial: 6A52; + 317440 bytes downloaded
[23:16:44] Initial: B478; + 327680 bytes downloaded
[23:16:44] Initial: CF8A; + 337920 bytes downloaded
[23:16:44] Initial: 8407; + 348160 bytes downloaded
[23:16:44] Initial: 2246; + 358400 bytes downloaded
[23:16:44] Initial: 1C69; + 368640 bytes downloaded
[23:16:44] Initial: 1287; + 378880 bytes downloaded
[23:16:44] Initial: 19B3; + 389120 bytes downloaded
[23:16:44] Initial: 1AD1; + 399360 bytes downloaded
[23:16:44] Initial: 5791; + 409600 bytes downloaded
[23:16:44] Initial: 76C5; + 419840 bytes downloaded
[23:16:44] Initial: 9B77; + 430080 bytes downloaded
[23:16:44] Initial: E82F; + 440320 bytes downloaded
[23:16:44] Initial: D0D3; + 450560 bytes downloaded
[23:16:44] Initial: 0F5E; + 460800 bytes downloaded
[23:16:44] Initial: D743; + 471040 bytes downloaded
[23:16:44] Initial: 0B7C; + 481280 bytes downloaded
[23:16:44] Initial: FAFD; + 491520 bytes downloaded
[23:16:44] Initial: 0E14; + 501760 bytes downloaded
[23:16:44] Initial: 4048; + 512000 bytes downloaded
[23:16:44] Initial: 21A5; + 522240 bytes downloaded
[23:16:44] Initial: C1A5; + 532480 bytes downloaded
[23:16:44] Initial: F716; + 542720 bytes downloaded
[23:16:44] Initial: DD98; + 552960 bytes downloaded
[23:16:44] Initial: 9F7B; + 563200 bytes downloaded
[23:16:44] Initial: 1CC0; + 573440 bytes downloaded
[23:16:44] Initial: 4D37; + 583680 bytes downloaded
[23:16:44] Initial: 222A; + 593920 bytes downloaded
[23:16:44] Initial: 8E33; + 604160 bytes downloaded
[23:16:44] Initial: D3C9; + 614400 bytes downloaded
[23:16:44] Initial: 9821; + 624640 bytes downloaded
[23:16:44] Initial: 236E; + 634880 bytes downloaded
[23:16:44] Initial: 1A7A; + 645120 bytes downloaded
[23:16:44] Initial: 6D64; + 655360 bytes downloaded
[23:16:44] Initial: 4ADC; + 665600 bytes downloaded
[23:16:44] Initial: 3854; + 675840 bytes downloaded
[23:16:44] Initial: CB5C; + 686080 bytes downloaded
[23:16:44] Initial: 2A88; + 696320 bytes downloaded
[23:16:44] Initial: 1199; + 706560 bytes downloaded
[23:16:44] Initial: 0512; + 716800 bytes downloaded
[23:16:44] Initial: 316E; + 727040 bytes downloaded
[23:16:44] Initial: D89D; + 737280 bytes downloaded
[23:16:44] Initial: E6A3; + 747520 bytes downloaded
[23:16:44] Initial: B488; + 757760 bytes downloaded
[23:16:44] Initial: BAFD; + 768000 bytes downloaded
[23:16:44] Initial: 34A0; + 778240 bytes downloaded
[23:16:44] Initial: DD6C; + 788480 bytes downloaded
[23:16:44] Initial: D2E9; + 789667 bytes downloaded
[23:16:44] Verifying core Core_a1.fah...
[23:16:44] Signature is VALID
[23:16:44]
[23:16:44] Trying to unzip core FahCore_a1.exe
[23:16:45] Decompressed FahCore_a1.exe (2035712 bytes) successfully
[23:16:50] + Core successfully engaged
[23:16:55]
[23:16:55] + Processing work unit
[23:16:55] Work type a1 not eligible for variable processors
[23:16:55] Core required: FahCore_a1.exe
[23:16:55] Core found.
[23:16:55] Using generic mpiexec calls
[23:16:55] Working on queue slot 01 [December 17 23:16:55 UTC]
[23:16:55] + Working ...
[23:16:55] - Calling 'mpiexec -np 4 -channel auto -host 127.0.0.1 FahCore_a1.exe
-dir work/ -suffix 01 -checkpoint 5 -forceasm -verbose -lifeline 3332 -version
622'

[23:16:55]
[23:16:55] *------------------------------*
[23:16:55] Folding@Home Gromacs SMP Core
[23:16:55] Version 1.74 (March 10, 2007)
[23:16:55]
[23:16:55] Preparing to commence simulation
[23:16:55] - Assembly optimizations manually forced on.
[23:16:55] - Not checking prior termination.
[23:16:57] - Expanded 2427788 -> 12894245 (decompressed 531.1 percent)
[23:16:58] - Starting from initial work packet
[23:16:58]
[23:16:58] Project: 2653 (Run 32, Clone 19, Gen 95)
[23:16:58]
[23:16:58] Assembly optimizations on if available.
[23:16:58] Entering M.D.
[23:17:04] Rejecting checkpoint
[23:17:05] OPC
[23:17:05] Writing local files
[23:17:05] Extra SSE boost OK.
[23:17:05]
[23:17:06] Extra SSE boost OK.
[23:17:06] Writing local files
[23:17:06] Completed 0 out of 500000 steps (0 percent)

i thought this was supposed to run in the taskbar

you got it. that's an smp unit. the smp client only runs in a cmd window...
there is a guide around here somewhere for configuring it as a service, but I've always used the cmd window method.

ok but i cant close this cause it will close it out completely correct, ok what can i use to see how long its going to take

ok i noticed that you got two windows open my i ask why ,cause i only got one open

you only need one on a phenom..

see above posts fahmon monitoring http://www.xtremesystems.org/forums/...&postcount=607

i dont think i have anything for the GPU

i have 4850 512 MB (Crossfire)

fahmon is a smp monitoring tool too.

so it not going to show my GPU's ?

Confused

fahmon will monitor all your clients, all you have to do is point it to the dir where the fahlog.txt file is located. why dont you install it and try... you can read the post I linked, it explains it pretty well.

Quote:

---

Originally Posted by MikeB12

see above posts fahmon monitoring http://www.xtremesystems.org/forums/...&postcount=607

---

ok i get something like this in the message window


[17/12/08 - 19:12:02] Finished Reloading cpu
[17/12/08 - 19:17:02] Reloading cpu
[17/12/08 - 19:17:02] cpu is on frame 4
[17/12/08 - 19:17:02] cpu [ETA] | avg=00:11:06 | ref=00:10:44 | adj=00:06:18 | left=17:04:06
[17/12/08 - 19:17:02] Finished Reloading cpu
[17/12/08 - 19:19:07] Reloading GPU
[17/12/08 - 19:19:07] X Error while reading C:\Documents and Settings\Roger\Application Data\Folding@home-gpu\unitinfo.txt!
[17/12/08 - 19:19:07] GPU is stopped (The line "Folding@Home Client Shutdown." was found)
[17/12/08 - 19:19:07] Finished Reloading GPU
[17/12/08 - 19:22:02] Reloading cpu
[17/12/08 - 19:22:02] cpu is on frame 5
[17/12/08 - 19:22:02] cpu [ETA] | avg=00:10:55 | ref=00:10:56 | adj=00:00:43 | left=17:17:57
[17/12/08 - 19:22:02] Finished Reloading cpu
[17/12/08 - 19:22:51] Reloading GPU
[17/12/08 - 19:22:51] X Error while reading C:\Documents and Settings\Roger\Application Data\Folding@home-gpu\unitinfo.txt!
[17/12/08 - 19:22:51] GPU has an unknown state (Unable to find a complete frame)
[17/12/08 - 19:22:51] Cannot determine frame number for GPU (this isn't a problem)
[17/12/08 - 19:22:51] GPU [ETA] | avg=N/A | last=N/A
[17/12/08 - 19:22:51] Finished Reloading GPU

why don't yall start a new thread for whatever client you're having an issue with. post your issue with your log.

then you can localize all your responses in the single thread for that specific client and problem. this bouncing around in this thread from different issues with different clients and users is not very clear.

Completely reinstalled both smp's and followed the tutorial exactly and still the same thing , only one smp running, but i was able to get the FAHMON to monitor the gpu.

maybe an issue with these new core i7's

I'm clueless how to fix this.

Quote:

---

Originally Posted by MikeB12

ok, the ppd on that smp looks like a single core unit.
stop the clients, delete the work folder and queue.dat file from your smp install directories to get rid of the single core untis.
you need to follow that walkthru to the letter, including adding the flags in the shortcut. http://www.xtremesystems.org/forums/...d.php?t=199283

make sure you have typed the smp and local flags with a space between in the desktop shortcuts,

Code:

---

Win32-x86.exe -smp -local

---

this tells the client to run in smp mode, instead of single core mode.. then when you launch the smp clients you should have 8 instances of fahcore_a1 in taskmgr. 4 per client.

---

i7 should run it fine.. I know 2 users here on them and have run dual smp's in windows and linux..

make a new thread and post the bigginning 50 lines or so of your fahlog.txt files... and maybe a screenshot of taskmgr sorted by cpu column..

I'm sure it's doable, we just need to figure out what's going on. and it's better to get it in it's own thread, instead of buried in here with a bunch of other posts and problems and hellos in the new user welcome thread.

Hello everyone! :wave:

Started folding now that I got a somewhat decent card again(Asus 4830) , while I wait for my sapphire 4870 rma :shakes:

I got fah tray client and fahmon setup, is there something else I need to know e.g. optimized clients, tweaks or startup parameters?

According to fahmon my ppd is about 2k-2,1k(proj 5733), clocking from stock 575 to 700 only changes ppd by about 200-300.

edit: seems to have gone up now 2,5k-2,7k :)

I am not familiar with AMD but they do get less points and points are down for everything right now so that is probably right.

I'm thinking of trying to build up an asrock p35 board with a spare q6600. I only have 2gig spare just now but will get 4 later if need be. The Plan is to move 2 gx2's to it and have a go at learning some way of using these a2 units. So whilst I have work to do trying to overclock on a 4core twins board, the bit I'm less sure about is which operating system to use. so folders what do you think? get another copy of windows (which one?) and then learn what? VMware and a linux within? or a Linux (which?) and ...well frankly I'm not at all clear what else I would need to do to run 4 GPU's.

I want my x38 back for day to day use and to try out some new toys including some watercooling trials.

I'd appreciate your views guys

Well if your not terrified by a terminal window for some setup and configuring and having to manually install some of the required software you could get more point out of a Linux setup and currently I would recommend Mepis 8 beta 6 64bit. Its a Debian (Lenny) base distribution which uses the KDE desktop and seems to work very well and seems to be kept up to date very well. Most of the setup for the SMP and GPU2 client using wine for Ubuntu will be the same and they also have a easy script to install the Nvidia video drivers which were a pain previously. The a2 smp core is a thing of beauty ppd wise on 3.6ghz E8400 C2duo with 2 gig of ram a 2665 a1 wu 1920pts will give me a ppd about 2.1kppd where a 2669 a2 wu 1920pts will give me a ppd about 3.2kppd. Its been shown that the GPU2 client running with wine is pretty much equal on ppd to a GPU2 client running in windows. Also it does not take a core to feed the GPU2 client and no dummy plug is needed for multiple cards and GPU2 clients on the same machine. Another thing going native Linux vs VMware will require less ram as well so your two gig should be good.

Of course I'm terrified...I'm pretty useless with windows but I muddle through by following the knowledge of others...so I'm hoping someone like yourself will help me through this when the time comes. Anyway, first problem solved ...no windows to buy!!!

Next problem: I've got 8 of these: http://www.delta.com.tw/product/ps/s...PS-600DB_A.pdf They can be included in two's or three's in this build if someone can help me connect a normal 24 way connector (which I don't even know the pinouts for) so I guess that's what I need first...can anyone link me to info on 24 way connector (which pin does what) to see if I can match the "signal" pins shown above

Edit after many times looking I found a diagram...now to see what matches

Quote:

---

Originally Posted by Shadowtester

Well if your not terrified by a terminal window for some setup and configuring and having to manually install some of the required software you could get more point out of a Linux setup and currently I would recommend Mepis 8 beta 6 64bit. Its a Debian (Lenny) base distribution which uses the KDE desktop and seems to work very well and seems to be kept up to date very well. Most of the setup for the SMP and GPU2 client using wine for Ubuntu will be the same and they also have a easy script to install the Nvidia video drivers which were a pain previously. The a2 smp core is a thing of beauty ppd wise on 3.6ghz E8400 C2duo with 2 gig of ram a 2665 a1 wu 1920pts will give me a ppd about 2.1kppd where a 2669 a2 wu 1920pts will give me a ppd about 3.2kppd. Its been shown that the GPU2 client running with wine is pretty much equal on ppd to a GPU2 client running in windows. Also it does not take a core to feed the GPU2 client and no dummy plug is needed for multiple cards and GPU2 clients on the same machine. Another thing going native Linux vs VMware will require less ram as well so your two gig should be good.

---

I tried installing mepis twice in Vmware. I let grub install in the default location and the system wouldn't boot. It said "Kernel Panic" Yada.Yada.Yada. Should I have installed grub to "root" instead?

Yes grub has to be installed in the boot partition to work if you installed it in the VMware partition it will not work very well Grub is a great boot manager I use it to do a quadruple boot on my Shuttle I need to reduce that some since I never use one of the os's and I replaced a second. I have Grub managing Mepis8 beta 6 64 bit, Kubuntu 8.04 64 bit, Kubuntu 7.10 32 bit, and Windows XP 32 bit. I need to remove the two Kubuntu installs after I get all the info I want to save off them to free up some hdd space.

And OldChap I would be glad to try to help you but over the holidays my time around the forums will be very limited at best and depending on the winter weather I could be really busy at work as well since I am a Auto Body repair and painting tech and winter can lead to some very long days at work with bad weather.

Thanks shadow, beggars can't be choosers and besides I have the physical build to deal with first...then the overclock that should take most of the holiday as case modding is a first for me too.

Maybe when I'm ready and if you have the time......

No problem I will be watching for your cries for help :)

On a related side note Mepis Linux just released their first Mepis8 RC1 for both 32 and 64 bit so I guess I will get to upgrade from the beta over the holiday I would expect a final release in early January.

I'm quite interested in trying this myself. My assumption is that VMware would be unnecessary if installing native Mepis...then just run multiple instances of the SMP and GPU clients?

Quote:

---

Originally Posted by coo-coo-clocker

I'm quite interested in trying this myself. My assumption is that VMware would be unnecessary if installing native Mepis...then just run multiple instances of the SMP and GPU clients?

---

That is true if you do not mind setting up wine to run the GPU2 client its not really that hard to do the hardest part is getting the Nvidia driver installed and there is a script you can use with Mepis to do that for you with minimal problems.

im also wanting to trying out wine i got crossfire on my ubuntu box doing nothing at all does wine support the ati gpu2 client ,

Quote:

---

Originally Posted by shadowwind

im also wanting to trying out wine i got crossfire on my ubuntu box doing nothing at all does wine support the ati gpu2 client ,

---

As far as I know not currently they were working on it over on the Stanford Folding Forum but I do not think they got it working yet you might check out that forum in the GPU ati sub forum for more info on the status of using wine with the ati card and gpu2 client.

i found a closed beta looks like its going to be released after the holidays so things are looking good for ati and wine.
to every one have a merry xmass and may god be with all of you peace.... from zozo101

is there a benchmark utility for F@H I plan on setting up several higher end folding machines and would like to see what little tweaks help out :) and also is there a single setup guide for cpu and gpu clients?

fahmon is what we use to get a fah benchmark PPD (points per day) reading on our clients http://fahmon.net/
there are several guides depending on what hardware and OS you're using. fah also has guides linked next to their dl's.

if you need some guidance, post a new thread with your hardware/os, and we can help you get setup on a client combo that will produce max points; it'll take the guesswork out of the process for you. but it's best to post a new thread instead of burying it in here in a welcome thread..
************************************************** *****************************
for example:
New thread: I need help on a client setup for this machine, here's my hardware/os?
OS=Vista
CPU=Q6600@3.33ghz
GPU=8800gt

reply:
1. run dual smp for your Q6600 per these instructions http://www.xtremesystems.org/forums/...d.php?t=199283
2. and the gpu systray for your gpu http://folding.stanford.edu/English/DownloadWinOther (first one on the pg)
after you get these setup post back with questions or problems
************************************************** ****************************
here is what fahmon data looks like

Up till now I've only ever used the console version of GPU .....is there any advantage in using systray? (I'm currently re-installing everything)

Quote:

---

Originally Posted by MikeB12

fahmon is what we use to get a fah benchmark PPD (points per day) reading on our clients http://fahmon.net/
there are several guides depending on what hardware and OS you're using. fah also has guides linked next to their dl's.

if you need some guidance, post a new thread with your hardware/os, and we can help you get setup on a client combo that will produce max points; it'll take the guesswork out of the process for you. but it's best to post a new thread instead of burying it in here in a welcome thread..
************************************************** *****************************
for example:
New thread: I need help on a client setup for this machine, here's my hardware/os?
OS=Vista
CPU=Q6600@3.33ghz
GPU=8800gt

reply:
1. run dual smp for your Q6600 per these instructions http://www.xtremesystems.org/forums/...d.php?t=199283
2. and the gpu systray for your gpu http://folding.stanford.edu/English/DownloadWinOther (first one on the pg)
after you get these setup post back with questions or problems
************************************************** ****************************
here is what fahmon data looks like

---

How much ppd do you lose if you use xp instead of vista?
My idea is to run dual SMP and a gpu client.
but i heard that in xp you need to free 1 core to feed the gpu client. But what about vista?
How big is the loss in ppd?

if you have to lock smp down to 3 cores, then dual smp is a no go. you'll lose about 1-2k ppd...
last time I used xp I had to run 1 smp with no affinity changer, lock it to 3 cores and run one gpu.
that's why I switched to vista last summer... I can run dual smp with AC on all 4 cores, and still run a gpu without it getting bottlenecked.. it only uses about 3-5% cpu in vista. 25% in xp.
but if I tried to run smp on 4 cores and a gpu in xp, then smp only allowed the gpu to have about 12% cpu and it choked the gpu ppd.
one of the xp guys needs to input here though, I haven't used xp in 6 months. and I heard nvidia came out with a new xp driver that allows gpu to turn full ppd without hogging a core to itself.

you'll have to try for your self to see... or else one of the xp guys here can chime in.
when I tried to run smp with gpu in xp, gpu turned low ppd unless you fed it an entire core.

but it's been 6 months since I tried, all the ppd you see in my sig is on vista.
(3 Q6600's below in Vista on dual smp w/AC and one gpu each- machines labeled x4,x5,x6)

Quote:

---

Originally Posted by MikeB12

if you have to lock smp down to 3 cores, then dual smp is a no go. you'll lose about 1-2k ppd...
last time I used xp I had to run 1 smp with no affinity changer, lock it to 3 cores and run one gpu.
that's why I switched to vista last summer... I can run dual smp with AC on all 4 cores, and still run a gpu without it getting bottlenecked.. it only uses about 3-5% cpu in vista. 25% in xp.
but if I tried to run smp on 4 cores and a gpu in xp, then smp only allowed the gpu to have about 12% cpu and it choked the gpu ppd.
one of the xp guys needs to input here though, I haven't used xp in 6 months. and I heard nvidia came out with a new xp driver that allows gpu to turn full ppd without hogging a core to itself.

you'll have to try for your self to see... or else one of the xp guys here can chime in.
when I tried to run smp with gpu in xp, gpu turned low ppd unless you fed it an entire core.

but it's been 6 months since I tried, all the ppd you see in my sig is on vista.
(3 Q6600's below in Vista on dual smp w/AC and one gpu each- machines labeled x4,x5,x6)

---

thx
then vista will be :)

I have been running 2*smp and 3 GPU successfully using xp by using a combination of 180.60 drivers and WinAFC ....The 180.60 (and later?) drivers have reduced the cpu requirement to almost nil, The use of WinAFC optimises the use to get maximum ppd.

If you use older drivers eg.179.84 for your GPU you will probably get an xtra 2-300ppd but using later 180.60 whilst reducing GPU increased smp by maybe 1000 when optimised

I paused the progress of my folding and adjusted my GPU's numbers. I then resumed the progress of the folding, and it started folding test proteins. Was changing the GPU's clocks in mid fold a bad thing?

Not Really, it will finish where it stopped if it didn't took you 2 days , then the WU's are no good but if you just stopped the folding for some reason and started it later , it works without issues

I've changed my gpu clocks during folding and it's fine.. shader is the only thing that matters, core and ram oc has no effect.
when you get too high on the shader it will eue, but that's unavodiable.

that's actually how I find my max shader oc.
1. start the client, fahmon, and precision tool or riva tuner.
2. bump the shader up at one hop at a time and let it fold for 5-10 minutes while watching/refreshing fahmon.
3. when it eue's, back it down one or two hops and that's the max shader folding oc.

(a hop is about 45mhz on my nividia cards)

I know this is probably a dumb question but is there a gpu console client? I hate running the systray for the pure fact it taxes the cpu so much.

Never mind my uneducated butt found it.

systray doesn't tax the CPU if it is kept in the tray. At least i didn't think it did :confused:
If it does maybe I should get off the systray clients :rolleyes:

Which OS and driver are you using Pulsar?

Okay well issue is now I'm moving 3 9800GX2's over to an AMD machine running vista32. I only have one 9800GX2 in right now I set up the flags correctly but when I try to run the 2nd core, it says unstable machine and restarts the console. Could this be because I don't have a monitor hooked up to the 2nd dvi?

^^ what happens to mine when the desktop is not extended is it gets a popup message and won't start... but mine was eueing real bad on one client. and I finally figured out it pulled a ati core for some stupid reason.. maybe that's what yours did. try deleting the fahcore_11.exe file, work folder, and queue.dat... and see if it fixes it...

I only have one monitor, but it has 2 connections on the back, a vga and dvi. so I used those to extend the desktop. once it's extended, I can unplug the vga, but powersave has to be turned off so monitor is always on, or else that second card loses it's desktop extension overnight. and when I reboot, sometimes I need to plug the vga back in, or the second unconnected desktop is not extended again.

just some stuff that plagued me in the beginning of dual gpu folding.

I just decided to go back to winxp so I don't have to deal with this multiple monitor issue. I finally go to hook up my tri 9800GX2 but the brand new Corsair HX1000W and nothing so starting the RMA process with that. Does anyone know if the 9800GX2 throttles itself due to temps?

Hey guys,

quick question: I kind of feel bad having my 4870X2 idle all the time, but I can't find any GPU folding project except for this one - is there another? Anyway, as some might remember, a while back I tried to get FAH running on both my X2's cores. The results were not encouraging, to say the least. Got only 1 core running and the PPD was ridiculously low. Did something change in the meantime regarding PPD of ATI cards and the, well, "ease of setup" or would I still be wasting countless hours and 300W of power drain to get beat by a 9600GSO? :p:

Hi everybody. I was hoping i could get some help on my smp client. On my quad I can only get it to run on 2 cores. I noticed it was only getting as many ppd as my dual core clocked the same. When i looked in the log it showed it only running on core 0 and 1. ?? any help would be appreciated

Did you install deino? or tried changing the affinity to all four cores?

Set up a second instance to run on the other two cores :D Seriously that will give you the best ppd and if you use WinAFC for your affinity setting you could be talking almost 50% more.

One of the guys re-posted a (6?) step guide to setting this up but I don't remember where...anyone with a working brain help?

I cant install deino for some reason. If someone could get that guide it would be awsome. I have tried setting up 2 instances of the client but it says to me there are no more cores. I have this machine dual booting vista and windows 7 beta and I think it is causing some problems with some programs installing correctly. When I ran the installer for the new nvidia driver from the vista desktop it would not install. It keep telling me that this OS was not supported, but when I moved the installer to the vista C: drive it installed correctly. Maybe the 7 install has taken "possession" of all the files on the drives it sees.

windows sees all 4 cores right?

When setting up clients it asks which projects you would like to download, its like small, normal, or large does that really have affect on how many ppd you earn?

At CPU I usually select normal and bug for GPU ... but the difference .. hmm didn't notice any

I'm just messing around with the new cuda 2.1's and I have noticed a 10C drop in load temps over 2.0, and also a 5% decrease in power consumption. Can anyone confirm this?

Quote:

---

Originally Posted by OldChap

Set up a second instance to run on the other two cores :D Seriously that will give you the best ppd and if you use WinAFC for your affinity setting you could be talking almost 50% more.

One of the guys re-posted a (6?) step guide to setting this up but I don't remember where...anyone with a working brain help?

---

Lowrez, This was it:
http://www.xtremesystems.org/Forums/...d.php?t=199283

Quote:

---

Originally Posted by =[PULSAR]=

I'm just messing around with the new cuda 2.1's and I have noticed a 10C drop in load temps over 2.0, and also a 5% decrease in power consumption. Can anyone confirm this?

---

Without a drop in points? Do I get this right, you're running 180.60? I could use a drop of 10°C. :yepp:

They are 181.2, Before I had system stability issues with Cuda 2.0 trying to run the card at stocks speeds. Now, running cuda 2.1 I have the core/shaders @ 700/1750 and running great so far. I just wish there was a way to OC just the shaders, but I haven't found a way yet while running 3 cards.

Quote:

---

Originally Posted by =[PULSAR]=

They are 181.2, Before I had system stability issues with Cuda 2.0 trying to run the card at stocks speeds. Now, running cuda 2.1 I have the core/shaders @ 700/1750 and running great so far. I just wish there was a way to OC just the shaders, but I haven't found a way yet while running 3 cards.

---

Okay, 181.20 don't work for me. But for your problem, what cards are you running? You can usually unlink core and shaders, to get more out of them. It worked for my GSOs and for the 295 too. Are you clocking with RivaTuner?

I have 3x EVGA 9800GX2's I've tried using rivatuner but will only let me OC one card, so I've been having to use nTune. Maybe a bios update?

Quote:

---

Originally Posted by =[PULSAR]=

I have 3x EVGA 9800GX2's I've tried using rivatuner but will only let me OC one card, so I've been having to use nTune. Maybe a bios update?

---

That's usually a result of not having displays extended to the desktop via dummy plugs. Other than that a detect on reboot has always worked for me. :)

What he said. You need to have your desktop extended to at least the primary display port of each "card" per gx2, then RT will be able to detect it you can clock all you want.

Well now I know my Sunday project this week :D

What was your OS again? Just to know if you'll need dummies or not..

WinXP 32

Then you can forget about the dummies AFAIR. :)

yeah, still need to "extend" the desktop to make it work tho.

Switched my folding rig over to a DFI X48 because of the stability issues with my DFI 790FX-M2R board and I can't recognized my third card at all. So, back down to folding with just two cards :( Anyone here folding with three cards on a P35 board???

Sparky thanks for the advice folding my 9800GX2 @ 740/1900/2000

Just started folding again and have a question or 2. I am folding with a 4850 and their site recommended the system try version of gpu client. is this ok to run or would it be better to run a different client?

My system is currently a I7 920 on foxconn Bloodrage with 6gb Supertalent 1666 ram. Also folding with a GPU; I take it that you cannot run different instances per core since its should actually be using gpu instead, correct?

also running on my secondary system which has X1950XT, 9550, 4gigs ram, OC'd to 4ghz. I did not see a client that supports the older GPU's; can someone tell me if there is still a version that would support an X1950XT; or would it be better to fold off of the cpu, or get an upgraded card?

Thanks for any help or input

INFRNL

I run a systray one so I can easily turn it off when gaming. I don't think the X1950XT is supported. I'd put SMP on it for the CPU.

I am new here and decided to join the team and help.

Quote:

---

Originally Posted by mikey063

I am new here and decided to join the team and help.

---

Well done! :up:

Quote:

---

Originally Posted by mikey063

I am new here and decided to join the team and help.

---

:welcome:

Thanks for joining up :up: If you have any setup questions or technical difficulties let us know, would be glad to help.

Quote:

---

Originally Posted by mikey063

I am new here and decided to join the team and help.

---

Welcome aboard! :)

Yep, what he said!
You'll find us to be a supportive, if not lewd, crude, and occasionally obnoxious crowd - but always looking for a good time.
Plus, several of us love to hunt and play with guns!
Attachment 94166

And we like girls in bikinis. Spring break race coming up!!!
http://www.spadre.com/images/springbreak.jpg

*cries* Guys ... start giving penalties , usually when I start with it Dak gets a heart attack from overdrooling

Has anyone had issues trying to run multiple nvidia cards on crossfire chipsets?

You just can't run them in SLI is all. They still work individually, which is how they need to be when folding anyway.

Mak2000, I see your penalty.

I think I have a better one.
http://aycu24.webshots.com/image/496...6188060_rs.jpg

Quote:

---

Originally Posted by =[PULSAR]=

Has anyone had issues trying to run multiple nvidia cards on crossfire chipsets?

---

Multiple Nvidia cards on a Cross-fire board work great for folding! :up:

Quote:

---

Originally Posted by WFO

Multiple Nvidia cards on a Cross-fire board work great for folding! :up:

---

I have 3 GSOs on an msi k9a2. Nuff said. :D

Maybe its just a problem with my DFI X48. Still waiting on my P35's back from DFI they have the slowest RMA department ever!!! I hope their reading this.

Just started folding for your team, quick question though. Do i need to submit something manually after i stop folding or does it submit everything by itself?

Quote:

---

Originally Posted by Mpetrosino

Just started folding for your team, quick question though. Do i need to submit something manually after i stop folding or does it submit everything by itself?

---

Once the WU is done it automatically uploads the results and then downloads the new WU. BTW thanks for joining the team :)