• 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.

F@H issues

A

Andross182

Gawd
Joined
Sep 20, 2005
Messages
527
I figured I would give the new F@H client a try and start folding for the Horde...

I followed this thread: http://hardforum.com/showthread.php?t=1550668

downloaded the GUI, setup the client and joined team 33...started the SMP client and all I get is a slew of errors....

All I get is an endless stream of...

[14:36:22] + Processing work unit
[14:36:22] Core required: FahCore_a3.exe
[14:36:22] Core not found.
[14:36:22] - Core is not present or corrupted.
[14:36:22] - Attempting to download new core...
[14:36:22] + Downloading new core: FahCore_a3.exe
[14:36:22] - Error: HTTP GET returned error code 0
[14:36:22] + Error: Could not download core
[14:36:22] + Core download error (#2), waiting before retry...

[14:36:39] + Downloading new core: FahCore_a3.exe
[14:36:39] - Error: HTTP GET returned error code 0
[14:36:39] + Error: Could not download core
[14:36:39] + Core download error (#3), waiting before retry...

[14:36:50] + Downloading new core: FahCore_a3.exe
[14:36:51] - Error: HTTP GET returned error code 0
[14:36:51] + Error: Could not download core
[14:36:51] + Core download error (#4), waiting before retry...

Did I set up something wrong? I tried to dig around on google but havent really found a solution.
 
Try giving the shortcut to the SMP client admin rights.
 
Sounds like a local network/proxy/firewall issue to me.
 
The PC that I am trying to get this to run on is not running through any ISA server...my firewall is a netgear FVS318 but being that I am able to download the client I dont see how that would cause an issue, unless you download the cores from elseware...
 
Can you please add the -verbosity 9 flag so we have a little more log file detail?
 
used logmein to get to my home PC....set it up the exact same way and it works there...but the network is setup identically with the same firewalls. Now I am really confused.
 
Firewall in this location had keyword blocking turned on, turned it off and it works...thanks gents :)
 
Andross182 said:
Firewall in this location had keyword blocking turned on, turned it off and it works...thanks gents :)
Click to expand...
Told ya so. :p Sorry, couldn't resist.. glad you worked it out. 99% of the time when you see an HTTP error in the log, it's a proxy/firewall issue.
 
I actually do have one more question....How much power draw does folding take? As the CPU and GPU are running at 100% load...
 
Andross182 said:
I actually do have one more question....How much power draw does folding take? As the CPU and GPU are running at 100% load...
Click to expand...
This depends very much on CPU/GPU architecture as some processors are more power hungry than others. Any overclocks will also factor into this. Tell us a little bit about what you are running.
 
Well I have the machine in my sig at home (will be using GPU/CPU folding), but then I have my work desktop which is an i7-860 and a AMD 3450 video card.

Also have two home servers. One running a Q9300 @ 2.5ghz and another running E6850 @ 3ghz.

Gonna run it by my boss to see if we can use our servers at work to fold as well, as they sit idle most of the time. :)
 
its hard to say for certain what the system will use but seeing as how the processor is stock probably in the 200-250w range with the 5870 folding.

the i7 860 probably 200w tops, the 2 home servers will definitely be under 200w.

the key thing is the psu though efficiency is very important with folding. if you plan on doing it a long time 80 plus silver and gold psu's are worth the extra money in the long run.
 
my PSUs are corsair 620hx and 520hx units which are 80 plus certified...dont know if they are bronze/silver/gold rated though
 
Andross182 said:
Well I have the machine in my sig at home (will be using GPU/CPU folding), but then I have my work desktop which is an i7-860 and a AMD 3450 video card.

Also have two home servers. One running a Q9300 @ 2.5ghz and another running E6850 @ 3ghz.

Gonna run it by my boss to see if we can use our servers at work to fold as well, as they sit idle most of the time. :)
Click to expand...

3450 - don't bother, you could fold on the v6 client, but you won't get enough points to offset the loss of points from the CPU. this card is non compatable with the v7 client.

the i7-860 - not sure the points, but at full load it will add around 150-200 watts. this is definitely worth your trouble. Set it up to fold SMP WUs, fold your 10 and get this thing running on the bigadv. overclocking it will help but either way, it should be just fine like this.

i7-920 - 130 watts (actually the TDP for this processor is 130, and w/o ocing folding will take you right below TDP), again, this cpu will do fine on SMP then even better on bigadv, and any ocing will make this even better even 20x180 for 3.6 people were seeing better than most video cards for less watts from the wall.

5870 - definitely worth folding - sort of. for the time being using the v7 client these crank out a good number of points at their TDP which is somewhere in the high 200-300 watt range. but to do this they need an entire CPU core to themselves. this will kill the points your 920 can make. So, I'd hold off on this card until the beta WUs pass and cpu usage for ATI goes down.

I'm not as familiar with the server hardware.
 
Last edited:
corsairs are good solid power supplies. They do not have the best efficiency, but they are solid. Many folders use corsair because they just don't fail and you can load them to within an inch of their life and they don't sag on voltage.

their efficiency should be in the mid 80s.

IMHO, they are perfect for the job, decent efficiency and high reliability
 
jfb9301 said:
3450 - don't bother, you could fold on the v6 client, but you won't get enough points to offset the loss of points from the CPU. this card is non compatable with the v7 client.

the i7-860 - not sure the points, but at full load it will add around 150-200 watts. this is definitely worth your trouble. Set it up to fold SMP WUs, fold your 10 and get this thing running on the bigadv. overclocking it will help but either way, it should be just fine like this.

i7-920 - 130 watts (actually the TDP for this processor is 130, and w/o ocing folding will take you right below TDP), again, this cpu will do fine on SMP then even better on bigadv, and any ocing will make this even better even 20x180 for 3.6 people were seeing better than most video cards for less watts from the wall.

5870 - definitely worth folding - sort of. for the time being using the v7 client these crank out a good number of points at their TDP which is somewhere in the high 200-300 watt range. but to do this they need an entire CPU core to themselves. this will kill the points your 920 can make. So, I'd hold off on this card until the beta WUs pass and cpu usage for ATI goes down.

I'm not as familiar with the server hardware.
Click to expand...

the only "client" I am using is the FAH GPU Tracker v2 that was posted in the thread. Is the v7 client a better choice? Do all ATI GPUs take a CPU core to run? If so ill shut down my 3450 and 5870.

I cant OC the i7-860 as it is in a Dell machine (work desktop) and right now with the 3450 folding its PPD is listed at just under 5k.

I can OC my i7-920 back up to 3.6ghz I suppose...my memory was giving me a few issues so I had to RMA it and just havent re-overclocked since.

Is there an overall tool I can use to track my PPD across all the machines I install the client on?

thanks for all your help guys, its time I put my spare machines to work! :)
 
The Dell should get you at least 10k ppd, depending on what you use the machine for. Mine does and I'm using it most of the day, at work.
 
Andross182 said:
the only "client" I am using is the FAH GPU Tracker v2 that was posted in the thread. Is the v7 client a better choice? Do all ATI GPUs take a CPU core to run? If so ill shut down my 3450 and 5870.

I cant OC the i7-860 as it is in a Dell machine (work desktop) and right now with the 3450 folding its PPD is listed at just under 5k.

I can OC my i7-920 back up to 3.6ghz I suppose...my memory was giving me a few issues so I had to RMA it and just havent re-overclocked since.

Is there an overall tool I can use to track my PPD across all the machines I install the client on?

thanks for all your help guys, its time I put my spare machines to work! :)
Click to expand...

for now, as far as I know, all ATI WUs take an entire thread to themselves. So on an i7 you should have 4 cores for a total of 8 threads. So when the SMP unit tries to use the 8th thread, it constantly has to wait on the ATI WU to give up process time. This slows things down a lot. Some people like me combat this by limiting the number of threads used by the SMP client (I use 10 instead of 12) but for the 3450 you cannot gain enough PPD off the 3450 to offset the loss by running 7 threads. Absolutely, shut down the 3450 and the PPD of that 860 should improve to something better than 5000 PPD.

HFM.net will track points for v6 clients but not v7

for the adventurous, v7 will unlock the potential of that 5870, but it will still cost an entire proccess thread to do it. on this one, the gain of the 5870 will more than offset the loss because under v7 it should make over 8000 PPD which will be more than you loose by cutting the 920 to 7 threads smp.

stanford has said that later ATI WUs will reduce the CPU load and increase the PPD, so that 5870 will be an even bigger beast PPD wise.
 
Just got back from running around my Dell i7-860 box is 11025 PPD....
where as the R410 2x Xeon E5630s @ 2.53 ghz with 32 gigs of memory is only pulling 9k....seems a little off lol
 
The R410 should be higher than that....I'm running bigadv on one with the same processors and it gets between 30k on a slower WU to 49k on the quicker ones.

Regular smp was around 26k, if I remember correctly.

Forgot to add, this is in Windows Server 2008 R2.
 
yeah thats the same OS as my server box, well with SP1....the VMs on it dont even put the CPU over 10% load....any ideas for me to check on mplee?
 
I have SP1 running on it also.

I'd post up the log so we can see if anything is strange in it. Is it loaded up 100% when folding? I had some issues with it downclocking when we first got it...turns out the heatsinks weren't installed properly, ie weren't screwed down and just resting on the cpu's. I'm not quite sure how that passed QA before it shipped....
 
Post up the log, when you get a chance. There might be some clue there on what's going on with it.
 
this may be a noobish question, and if so, forgive me but what logfile exactly are ya looking for? or are there any special paramaters I need to set to log first?
 
You may be seeing less ppd if you haven't reached the bonus point threshold yet. After you turn in 10 successful SMP WUs, as long as you maintain an 80% successful WU completion percentage, you'll earn bonus points for turning WUs in before the deadline.
 
I hope thats the case...as my r310 server with an x3440 @ 2.53 ghz has about 1k more ppd then it.
 
Are you using the GPU tracker program to fold? If so you can right click the icon in the system tray to get to the log file.

I don't think the tracker or any other monitoring program knows whether you completed the 10 units or not...it just assumes that you did when calculating the ppd.
 
This is all I get when I right click. Yes, I am using the GPU tracker v2
rightclick.jpg
\
no option for a log file export
 
Last edited:
Try a path similar to this for the log file...I assume its folding@home-cpu instead of gpu though.

C:\Users\*username*\Appdata\folding@home-gpu\FAHlog.txt

I don't use the tracker, but I pulled it from the tracker thread(thanks Zero82z).
 
This what ya looking for?

--- Opening Log file [April 28 10:16:40 UTC]


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

Folding@Home Client Version 6.34

http://folding.stanford.edu

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

Launch directory: C:\Program Files (x86)\F@H\F@H\FAH GPU Tracker V2\SMP
Executable: C:\Program Files (x86)\F@H\F@H\FAH GPU Tracker V2\FAH.exe
Arguments: -oneunit -forceasm -smp

[10:16:40] - Ask before connecting: No
[10:16:40] - User name: Andross182 (Team 33)
[10:16:40] - User ID: 4A049A5B06510F57
[10:16:40] - Machine ID: 2
[10:16:40]
[10:16:40] Work directory not found. Creating...
[10:16:40] Could not open work queue, generating new queue...
[10:16:40] - Preparing to get new work unit...
[10:16:40] Cleaning up work directory
[10:16:40] + Attempting to get work packet
[10:16:40] Passkey found
[10:16:40] - Connecting to assignment server
[10:16:40] - Successful: assigned to (128.143.199.96).
[10:16:40] + News From Folding@Home: Welcome to Folding@Home
[10:16:41] Loaded queue successfully.
[10:16:53] + Closed connections
[10:16:53]
[10:16:53] + Processing work unit
[10:16:53] Core required: FahCore_a3.exe
[10:16:53] Core found.
[10:16:53] Working on queue slot 01 [April 28 10:16:53 UTC]
[10:16:53] + Working ...
[10:16:53]
[10:16:53] *------------------------------*
[10:16:53] Folding@Home Gromacs SMP Core
[10:16:53] Version 2.27 (Dec. 15, 2010)
[10:16:53]
[10:16:53] Preparing to commence simulation
[10:16:53] - Assembly optimizations manually forced on.
[10:16:53] - Not checking prior termination.
[10:16:54] - Expanded 1770336 -> 1965868 (decompressed 111.0 percent)
[10:16:54] Called DecompressByteArray: compressed_data_size=1770336 data_size=1965868, decompressed_data_size=1965868 diff=0
[10:16:54] - Digital signature verified
[10:16:54]
[10:16:54] Project: 7153 (Run 0, Clone 53, Gen 80)
[10:16:54]
[10:16:54] Assembly optimizations on if available.
[10:16:54] Entering M.D.
[10:17:00] Mapping NT from 16 to 16
[10:17:00] Completed 0 out of 500000 steps (0%)
[10:22:04] Completed 5000 out of 500000 steps (1%)
[10:27:23] Completed 10000 out of 500000 steps (2%)
[10:32:32] Completed 15000 out of 500000 steps (3%)
[10:37:21] Completed 20000 out of 500000 steps (4%)
[10:41:37] Completed 25000 out of 500000 steps (5%)
[10:45:52] Completed 30000 out of 500000 steps (6%)
[10:50:07] Completed 35000 out of 500000 steps (7%)
[10:54:41] Completed 40000 out of 500000 steps (8%)
[10:59:32] Completed 45000 out of 500000 steps (9%)
[11:04:15] Completed 50000 out of 500000 steps (10%)
[11:09:00] Completed 55000 out of 500000 steps (11%)
[11:14:05] Completed 60000 out of 500000 steps (12%)
[11:19:50] Completed 65000 out of 500000 steps (13%)
[11:25:42] Completed 70000 out of 500000 steps (14%)
[11:31:03] Completed 75000 out of 500000 steps (15%)
[11:36:03] Completed 80000 out of 500000 steps (16%)
[11:41:18] Completed 85000 out of 500000 steps (17%)
[11:46:27] Completed 90000 out of 500000 steps (18%)
[11:52:06] Completed 95000 out of 500000 steps (19%)
[11:57:18] Completed 100000 out of 500000 steps (20%)
[12:01:59] Completed 105000 out of 500000 steps (21%)
[12:06:38] Completed 110000 out of 500000 steps (22%)
[12:11:26] Completed 115000 out of 500000 steps (23%)
[12:16:28] Completed 120000 out of 500000 steps (24%)
[12:21:51] Completed 125000 out of 500000 steps (25%)
[12:27:14] Completed 130000 out of 500000 steps (26%)
[12:32:24] Completed 135000 out of 500000 steps (27%)
[12:37:19] Completed 140000 out of 500000 steps (28%)
[12:42:05] Completed 145000 out of 500000 steps (29%)
[12:47:08] Completed 150000 out of 500000 steps (30%)
[12:52:07] Completed 155000 out of 500000 steps (31%)
[12:57:17] Completed 160000 out of 500000 steps (32%)
[13:02:05] Completed 165000 out of 500000 steps (33%)
[13:07:09] Completed 170000 out of 500000 steps (34%)
[13:12:49] Completed 175000 out of 500000 steps (35%)
[13:18:02] Completed 180000 out of 500000 steps (36%)
[13:23:32] Completed 185000 out of 500000 steps (37%)
[13:28:43] Completed 190000 out of 500000 steps (38%)
[13:34:15] Completed 195000 out of 500000 steps (39%)
[13:40:05] Completed 200000 out of 500000 steps (40%)
[13:45:22] Completed 205000 out of 500000 steps (41%)
[13:50:45] Completed 210000 out of 500000 steps (42%)
[13:55:56] Completed 215000 out of 500000 steps (43%)
[14:00:54] Completed 220000 out of 500000 steps (44%)
[14:06:24] Completed 225000 out of 500000 steps (45%)
[14:12:17] Completed 230000 out of 500000 steps (46%)
[14:17:49] Completed 235000 out of 500000 steps (47%)
[14:23:11] Completed 240000 out of 500000 steps (48%)
[14:28:24] Completed 245000 out of 500000 steps (49%)
[14:33:06] Completed 250000 out of 500000 steps (50%)
[14:37:44] Completed 255000 out of 500000 steps (51%)
[14:42:19] Completed 260000 out of 500000 steps (52%)
[14:46:58] Completed 265000 out of 500000 steps (53%)
[14:51:47] Completed 270000 out of 500000 steps (54%)
[14:56:26] Completed 275000 out of 500000 steps (55%)
[15:01:07] Completed 280000 out of 500000 steps (56%)
[15:05:58] Completed 285000 out of 500000 steps (57%)
[15:10:54] Completed 290000 out of 500000 steps (58%)
[15:15:48] Completed 295000 out of 500000 steps (59%)
[15:21:25] Completed 300000 out of 500000 steps (60%)
[15:26:41] Completed 305000 out of 500000 steps (61%)
[15:31:56] Completed 310000 out of 500000 steps (62%)
[15:37:52] Completed 315000 out of 500000 steps (63%)
[15:43:49] Completed 320000 out of 500000 steps (64%)
[15:49:20] Completed 325000 out of 500000 steps (65%)
[15:54:44] Completed 330000 out of 500000 steps (66%)
[16:00:00] Completed 335000 out of 500000 steps (67%)
[16:05:02] Completed 340000 out of 500000 steps (68%)
[16:10:03] Completed 345000 out of 500000 steps (69%)
[16:14:59] Completed 350000 out of 500000 steps (70%)
[16:19:48] Completed 355000 out of 500000 steps (71%)
[16:24:35] Completed 360000 out of 500000 steps (72%)
[16:29:53] Completed 365000 out of 500000 steps (73%)
[16:35:28] Completed 370000 out of 500000 steps (74%)
[16:40:58] Completed 375000 out of 500000 steps (75%)
[16:46:19] Completed 380000 out of 500000 steps (76%)
[16:51:46] Completed 385000 out of 500000 steps (77%)
[16:57:11] Completed 390000 out of 500000 steps (78%)
[17:02:26] Completed 395000 out of 500000 steps (79%)
[17:07:33] Completed 400000 out of 500000 steps (80%)
 
Those times are just slighty quicker than a stock Q9650.

Is the server throttling back for some reason?
 
good question. I was going to try to delete the client and reset it up. The machine hosts VMs...that just run simple apps like quickbooks and what not...it literally has less the 5% cpu load alllll day.
 
The frame times are consistent enough that it seems like its just throttling constantly.

If you run cpuz, does it show the multiplier at 20? When ours was having the thottling issue, the multiplier would drop as soon as the server went under any kind of load(due to heat).
 
Intel Xeon E5630 -
Core Speed: 2527.1 mhz
Multiplier: 19x
Bus Speed 133 mhz
QPI link 2927 mhz
 
I'm out of ideas. This one is locked at 20, but I disabled the stuff in the bios for power savings. NUMA is disabled also, but I don't think it would make that much difference.

The server has 2 SAS drives mirrored with 8GB of RAM running SQL 2008 R2.
 
You must log in or register to reply here.
Back
Top