Go Back   Hardware Canucks > HARDWARE CANUCKS COMMUNITY > HardwareCanucks F@H Team

    
Reply
 
LinkBack Thread Tools Display Modes
  #291 (permalink)  
Old February 22, 2009, 06:23 AM
sswilson's Avatar
Moderator
F@H
 
Join Date: Dec 2006
Location: Moncton NB
Posts: 14,417

My System Specs

Default

I'm getting a "FAH Core Download error".... :(
__________________
MSI Z87I Gaming AC / i5 4670K / 2X 4G Gskill 1866 DDR3 / XFX XTR 750 / EVGA GTX 680 SC+ 2GB / Intel DC S3700 200G / random 160G Sata HDD
Inwin 904 / Swiftech MCP655-b / Alphacool NexXxos XT45 120 Rad / 2X Scythe GT AP-15 / EK Supreme HF / Dell UltraSharp U2412M

Asrock AM1H-ITX / AM1 Athlon 5350 / 2X4G Gskill PC3-14900 / Intel 6235 Wi-Fi / 90W Targus Power Brick / 320G Seagate Momentus / Mini-Box M350 / 1X 22" Dell IPS / 1X 22" HP
Reply With Quote
  #292 (permalink)  
Old February 22, 2009, 06:54 AM
LCB001's Avatar
Folding Captain
 
Join Date: Feb 2008
Location: Aylmer QC.
Posts: 1,774

My System Specs

Default

Quote:
Originally Posted by sswilson View Post
I'm getting a "FAH Core Download error".... :(
Checking the F@H forums, a new beta core Fah_14 got released by mistake and is messing up a lot of machine's. It can be fixed by ;

If you have a WU waiting to send, restart client, let it send it then;

Stopping the client
Deleting the work folder, queue.dat and unitinfo.
Restarting client

Great way to start the morning, thanks Stanford...
__________________
Folding For Team 54196

Reply With Quote
  #293 (permalink)  
Old February 22, 2009, 09:05 AM
Alwaysrun's Avatar
Allstar
F@H
 
Join Date: Sep 2008
Location: Qualicum Beach BC
Posts: 800

My System Specs

Default

I made a thread about shader OC'ing a month back and I just couldn't get mine over 1448 which was just one bump above stock for my card. Not until I bottomed out my core/mem was I finially able to get the shaders to 1620 and noticed an overall increase of ~450ppd. I think we all jumped onto the the shader bandwagon and some posts said they saw increases in production. So many things have happened since then for me (new drivers, new 64bit OS, better GPU cooling) it's hard to pinpoint what exactly happened to our shaders being a primary factor with folding. Upon my fresh OS install and newest drivers I OC'd my card just as before with the shaders pinned and the results were dismal compared and not untill I increased all three GPU components did I see results equal or surpassing previous ppd highs. I think someone alluded to a suspision that they make unannounced alterations to the clients FahCore_11.exe and maybe that has something to do with it but I couldn't tell as I did a fresh install. One thing I have noticed is a slight increase in CPU usage from 4-6% to 6-8% but that may just be because of my higher clocks.

I see why SKYMTL is leary about including folding stats in graphic card reviews as he said he'd have to retest every card again when new drivers came out and I see his point with diamond clarity now.

On another note I have a question about the SMP client. I used the Deino client before but now I have to use the MPICH because I'm using a 64bit OS even though I only have a dual core correct?

BTW grats everyone we hit our 2nd 200k day yesterday for the 2nd time in a week

*Edit; LCB they did WHAT! nooo don't mess with my stuff it's working awesome atm... whew I just scrolled through my log and I guess I missed that core update somehow..

Last edited by Alwaysrun; February 22, 2009 at 09:12 AM.
Reply With Quote
  #294 (permalink)  
Old February 22, 2009, 09:33 AM
chrisk's Avatar
Folding Captain
 
Join Date: Jul 2008
Location: GTA, Ontario
Posts: 7,401

My System Specs

Default

Quote:
Originally Posted by LCB001 View Post
The same thing is posted on all the different Folding forums I read...

There must be some magical cutoff numbers you can't go below, either that or your cards are just really really picky...
There is a limit on how much a core can be overclocked in relation to the shader it seems. According to this thread (eVGA GTX 260 (216SP) overclocking expectations - [H]ard|Forum), your core cannot go more than 1/2 the speed of the shaders. Perhaps there is something going on when you try to clock the shaders too high.

Could also have something to do with cooling. The idea (I thought) of underclocking the core was to decrease heat and stress on the chip, thus allowing the shaders to be adjusted higher. Perhaps the temp sensor is not reading the shader areas of the chip well, or some better tim will get better results. I also wonder when they bin the chips, do they stress the shaders as much as the core.
__________________
Fold for team #54196

Last edited by chrisk; February 22, 2009 at 10:13 AM. Reason: foggier than usual today
Reply With Quote
  #295 (permalink)  
Old February 22, 2009, 10:46 AM
LCB001's Avatar
Folding Captain
 
Join Date: Feb 2008
Location: Aylmer QC.
Posts: 1,774

My System Specs

Default

Quote:
Originally Posted by Alwaysrun View Post
I made a thread about shader OC'ing a month back and I just couldn't get mine over 1448 which was just one bump above stock for my card. Not until I bottomed out my core/mem was I finially able to get the shaders to 1620 and noticed an overall increase of ~450ppd. I think we all jumped onto the the shader bandwagon and some posts said they saw increases in production. So many things have happened since then for me (new drivers, new 64bit OS, better GPU cooling) it's hard to pinpoint what exactly happened to our shaders being a primary factor with folding. Upon my fresh OS install and newest drivers I OC'd my card just as before with the shaders pinned and the results were dismal compared and not untill I increased all three GPU components did I see results equal or surpassing previous ppd highs. I think someone alluded to a suspision that they make unannounced alterations to the clients FahCore_11.exe and maybe that has something to do with it but I couldn't tell as I did a fresh install. One thing I have noticed is a slight increase in CPU usage from 4-6% to 6-8% but that may just be because of my higher clocks.

I see why SKYMTL is leary about including folding stats in graphic card reviews as he said he'd have to retest every card again when new drivers came out and I see his point with diamond clarity now.

On another note I have a question about the SMP client. I used the Deino client before but now I have to use the MPICH because I'm using a 64bit OS even though I only have a dual core correct?

BTW grats everyone we hit our 2nd 200k day yesterday for the 2nd time in a week

*Edit; LCB they did WHAT! nooo don't mess with my stuff it's working awesome atm... whew I just scrolled through my log and I guess I missed that core update somehow..
You are correct, the Deino client only works with a 32bit OS, in your case the Mpich client is required.

The beta core release didn't hit everybody, luckily I only got it on one of mine. They are only supposed to be released to the beta testers until ready for general release. This is not the first time this has happened, it just means there is a new core in our future .

Quote:
Perhaps the temp sensor is not reading the shader areas of the chip well
This is something I've suspected for a while and have mentioned in previous posts, good airflow over the card seems to be the key regardless of what the temp. sensors say...
__________________
Folding For Team 54196

Reply With Quote
  #296 (permalink)  
Old February 22, 2009, 02:10 PM
Alwaysrun's Avatar
Allstar
F@H
 
Join Date: Sep 2008
Location: Qualicum Beach BC
Posts: 800

My System Specs

Default

Huston we're a GO. Everythings going swimingly with MPICH and the new/old CPU OC. I stuck to 3.85Ghz as it's the best power/heat sweet spot as I could find. Only thing odd is that when I set up MPICH I specified that it only use 92% CPU but when I first ran it and checked my system resources it said I was using 98% (This SMP client was all I had running not with GPU folding) I checked the console to see if I could see any text that said it was limiting to only 92% but I couldn't find any. wierd.

Name:  Fully OCd.jpg
Views: 69
Size:  151.3 KB

10,680ppd!
Reply With Quote
  #297 (permalink)  
Old February 22, 2009, 02:19 PM
chrisk's Avatar
Folding Captain
 
Join Date: Jul 2008
Location: GTA, Ontario
Posts: 7,401

My System Specs

Default

Tried the Deino install today on my folding rig and FaH crashes whenever I try to set the prefs for FaH. CPU is not OCed either, and neither is the ram. I'm gonna try the MPICH and see if something is up on that folding rig of mine.

Could really use the extra PPD as I just shipped off my Core 216 to Evga for the 'Step up' to a 285. Thats about 6000 PPD going poof for me for about a week or two.

Looks good Alwaysrun!
__________________
Fold for team #54196
Reply With Quote
  #298 (permalink)  
Old February 22, 2009, 02:29 PM
Alwaysrun's Avatar
Allstar
F@H
 
Join Date: Sep 2008
Location: Qualicum Beach BC
Posts: 800

My System Specs

Default

Thanks chrisk. I think it was LCB before that told me about selecting "Do not lock cores" and having the core priority set to "slightly higher" when using gpu and smp at the same time. Made perfect sense at the time but once again when I started up my GPU client I got those damn dirty NANs detected on GPU errors again. So I deselected both those options and now everything going fine. I'm thinking that those were the culprits before why I was getting the NANs errors...ohh well a good format, new OS, and a clean system was worth the effort and maybe I stumbled onto something the other users on the stanford forum were plagued with.
Reply With Quote
  #299 (permalink)  
Old February 22, 2009, 02:54 PM
chrisk's Avatar
Folding Captain
 
Join Date: Jul 2008
Location: GTA, Ontario
Posts: 7,401

My System Specs

Default

I think the folding rig has a stability issue when I try to run the cpu app. I don't get nans...I get "Application has encountered an error" error. Will figure it out.
__________________
Fold for team #54196
Reply With Quote
  #300 (permalink)  
Old February 22, 2009, 04:50 PM
3.0charlie's Avatar
3.0 "I kill SR2's" Charlie
F@H
 
Join Date: May 2007
Location: Laval, QC
Posts: 9,613

My System Specs

Default

I've got some better ppd, but there is still something fishy...

Both cards are 260s, consecutive serial numbers. Both are clocked exactly the same, but 260 #1 has a ppd of 5733 and 260 #2 has a ppd of 4747; both are folding a 511...
__________________
Hydro-Quebec is salivating...
Reply With Quote
Reply


Thread Tools
Display Modes