EVGA

New beta version of core 17 fixes GK104 - tested and confirmed

Page: 12 > Showing page 1 of 2
Author
Zagen30
SSC Member
  • Total Posts : 528
  • Reward points : 0
  • Joined: 2009/11/27 12:21:01
  • Location: Laguna Niguel, CA
  • Status: offline
  • Ribbons : 3
2014/04/15 10:34:08 (permalink)
*Update*

A few people have tested this and confirmed that this fixes the GK104 slowdown issue.  So, set your slot(s) to client-type beta if they aren't already, and while the slots are paused delete the beta core folder listed below to force the client to get the new version.
 
*Original*
 
Over on FF there's been chatter that the new beta version of core 17, 0.0.55, has fixed the low-PPD issue on GK104.  There's one report from a user with a 670 saying that it worked for him:
 
BleederExt64 is correct. 337.50 didn't correct the low PPD but it did run smoothly and without error on the "old" core.

Only when the v.55 core17 went beta did the PPD go back to "normal".

 
If someone could test this on their own GK104, that'd be useful.  You'd need to set the slot to client-type:beta if you haven't already.  None of the beta core 17 projects have this set as the minimum version, so you'd need to manually force an update by deleting the beta core 17 folder when the slot is paused (default C:\Users\<your name>\AppData\Roaming\FAHClient\cores\www.stanford.edu\~pande\Win32\AMD64\NVIDIA\Fermi\beta\Core_17.fah) and then restarting the slot.
 
I'd test this myself but I don't have a GK104.  I have been running this version for the last 5 days or so without issue on my 780s.  
post edited by Zagen30 - 2014/04/15 15:34:47


 
Core i7 3770K @4.3 GHz, 8 GB DDR3 @2133 MHz, 2xGTX 970 SSC @1379 MHz, Windows 7 Pro 64-bit    
 
Affiliate code: K4BJ4FHUX5
 
#1

44 Replies Related Threads

    boballee
    FTW Member
    • Total Posts : 1455
    • Reward points : 0
    • Joined: 2011/03/16 12:19:13
    • Location: Waiting for the Levee to Break at Bron-Yr-Aur
    • Status: offline
    • Ribbons : 8
    Re: New beta version of core 17 might fix GK104 - can someone test? 2014/04/15 11:00:36 (permalink)
    I'm running a 660 Ti and happy to help test this out.  I've been somewhat out of the loop for a bit, so I'm not really sure what, other than what you listed above, you want/need to see.  Here's my various specs:
     
    GTX 660 Ti FTW (only slot folding, set with a client-type: beta)
    Driver: 326.80
    Currently folding a 9101 @ 3:55 TPF
     



    #2
    Zagen30
    SSC Member
    • Total Posts : 528
    • Reward points : 0
    • Joined: 2009/11/27 12:21:01
    • Location: Laguna Niguel, CA
    • Status: offline
    • Ribbons : 3
    Re: New beta version of core 17 might fix GK104 - can someone test? 2014/04/15 11:13:26 (permalink)
    You'd need to update your drivers to something newer to see if the problem had been fixed.  326.80 was one of the last that didn't cause this issue.  I guess you might as well go for 337.50, the newest beta driver, or whatever the latest WHQL driver is.
     
    If, after updating the drivers and core, you're seeing frame times on par with what you're currently getting, that's strong evidence that they have fixed the issue.  If the bug is present, you should be looking at frame times in the 9 minute range for that project, assuming a roughly 70% drop in PPD that most others have reported.
    post edited by Zagen30 - 2014/04/15 11:22:11


     
    Core i7 3770K @4.3 GHz, 8 GB DDR3 @2133 MHz, 2xGTX 970 SSC @1379 MHz, Windows 7 Pro 64-bit    
     
    Affiliate code: K4BJ4FHUX5
     
    #3
    boballee
    FTW Member
    • Total Posts : 1455
    • Reward points : 0
    • Joined: 2011/03/16 12:19:13
    • Location: Waiting for the Levee to Break at Bron-Yr-Aur
    • Status: offline
    • Ribbons : 8
    Re: New beta version of core 17 might fix GK104 - can someone test? 2014/04/15 11:21:29 (permalink)
    Gotcha.  I'll update the drivers and see what happens.  



    #4
    boballee
    FTW Member
    • Total Posts : 1455
    • Reward points : 0
    • Joined: 2011/03/16 12:19:13
    • Location: Waiting for the Levee to Break at Bron-Yr-Aur
    • Status: offline
    • Ribbons : 8
    Re: New beta version of core 17 might fix GK104 - can someone test? 2014/04/15 11:55:29 (permalink)
    Alright - here's what I have:
    660 Ti FTW
    Driver version: 335.23
    Window's Vista Ultimate 64-bit
     
    I was about 60% through a 9101 WU, paused it, forced the core 17 update you listed above and restarted.  My current TPF is 3:57.  I'll keep an eye on it and report any changes.
     
    **EDIT: Now at 76% through the WU, I'm running a 3:54 TPF - so 'technically' lower than before.  
    post edited by boballee - 2014/04/15 12:48:34



    #5
    BadBertie
    SSC Member
    • Total Posts : 852
    • Reward points : 0
    • Joined: 2012/08/10 17:09:48
    • Location: France
    • Status: offline
    • Ribbons : 1
    Re: New beta version of core 17 might fix GK104 - can someone test? 2014/04/15 12:56:35 (permalink)
    I've got 3 x 670s running. I'll give it a crack too. Trouble is they're all stuck on core 15 projects atm.
    I'll try and shift things along a bit and let you know what happens

    edit
    3 x gtx 670
    Win7 64bit
    driver 335.23
    client type = beta
    one card is just about to finish core15 wu. Let's see if I can pick up something to test this with

    edit
    WU 9406
    Tpf 05:30 approx
    est ppd 71k
    never ran 9406 wus before so have nothing to compare directly.
    However have run many 9401s at steady tpf 09:49 and 71k ppd according to my hfm.net history.

    I'll let you know how it finishes.
    post edited by BadBertie - 2014/04/15 14:17:31

    KTM Super Duke R EVO
    #6
    theGryphon
    CLASSIFIED Member
    • Total Posts : 4069
    • Reward points : 0
    • Joined: 2009/03/04 17:27:46
    • Location: United States
    • Status: offline
    • Ribbons : 14
    Re: New beta version of core 17 might fix GK104 - can someone test? 2014/04/15 13:56:28 (permalink)
    Just tested it now, and it's good news it seems.
     
    GPU: GTX 760
    Clock: 1267MHz effective
    Driver: 335.23
    FahCore: Core 17 version 0.0.55
    WU: 9406 (19, 0, 19)
    TPF: ~6:00 (pretty variable)
    PPD: ~72K (inline with 327 series drivers)
     
    I had tried the 337.50 beta with my GTX 780Ti and it was very wonky. It caused tons of "Bad WU" and other errors. So, I'm hesitant to try again.
     
    #7
    Jackos
    Superclocked Member
    • Total Posts : 165
    • Reward points : 0
    • Joined: 2011/03/06 15:49:14
    • Status: offline
    • Ribbons : 1
    Re: New beta version of core 17 might fix GK104 - can someone test? 2014/04/15 14:50:46 (permalink)
    I confirm this is a proper fix. Folding with 337.50 at full speed on my two GTX760! :)

    Haswell RIG: MSI Z87 MPOWER; Intel Core i7-4770K; Corsair Vengeance Pro 2133Mhz CL11 16GB; EVGA GeForce GTX 980 ACX 2.0 w/ backplate; Samsung SSD PRO 850 256GB + Intel SSD 530 Series 180GB + Samsung SSD EVO 840 250GB; 2x WDC Black 3TB RAID0; CHIEFTEC BPS-750C2; Fractal Design Define R3 w/ USB3; Logitech MX Master + Wireless Illuminated Keyboard K800; Samsung PX2370; Windows 8.1 RTM Genuine

    #8
    bison88
    Superclocked Member
    • Total Posts : 136
    • Reward points : 0
    • Joined: 2013/10/28 17:44:21
    • Status: offline
    • Ribbons : 0
    Re: New beta version of core 17 might fix GK104 - can someone test? 2014/04/15 15:12:53 (permalink)
    This is great news!  Finally the gamers can update their fricken drivers without having to pay huge penalties to fold.
     
    Seems a 750/750 Ti fix is in-route next.  Although we wont see what they can truly produce according to VJ.
    #9
    Jackos
    Superclocked Member
    • Total Posts : 165
    • Reward points : 0
    • Joined: 2011/03/06 15:49:14
    • Status: offline
    • Ribbons : 1
    Re: New beta version of core 17 might fix GK104 - can someone test? 2014/04/15 15:15:22 (permalink)
    I'm also seeing a little bit of improvement in PPD comparing to 326.80. However, let it stabilize during the night...

    Haswell RIG: MSI Z87 MPOWER; Intel Core i7-4770K; Corsair Vengeance Pro 2133Mhz CL11 16GB; EVGA GeForce GTX 980 ACX 2.0 w/ backplate; Samsung SSD PRO 850 256GB + Intel SSD 530 Series 180GB + Samsung SSD EVO 840 250GB; 2x WDC Black 3TB RAID0; CHIEFTEC BPS-750C2; Fractal Design Define R3 w/ USB3; Logitech MX Master + Wireless Illuminated Keyboard K800; Samsung PX2370; Windows 8.1 RTM Genuine

    #10
    boballee
    FTW Member
    • Total Posts : 1455
    • Reward points : 0
    • Joined: 2011/03/16 12:19:13
    • Location: Waiting for the Levee to Break at Bron-Yr-Aur
    • Status: offline
    • Ribbons : 8
    Re: New beta version of core 17 might fix GK104 - can someone test? 2014/04/15 17:26:38 (permalink)
    Alright - here's another update:
     
    WU: 9406
    TPF: 4:10 (across 33 frames)
     
    I don't think I've ever had one of these, so I don't have anything to compare against - but if I keep this rate, it's putting me almost 124,000 PPD (which I haven't been anywhere close to in some time).  I'd say whatever fix they put in seems to be working.  



    #11
    Drazhar
    CLASSIFIED Member
    • Total Posts : 2370
    • Reward points : 0
    • Joined: 2008/12/07 02:14:51
    • Status: offline
    • Ribbons : 4
    Re: New beta version of core 17 might fix GK104 - can someone test? 2014/04/15 19:12:42 (permalink)
    This is certainly good news to see, even for those like me who I believe aren't affected.. (GTX 570 for a bit longer yet)


     
    #12
    bcavnaugh
    The Crunchinator
    • Total Posts : 38977
    • Reward points : 0
    • Joined: 2012/09/18 17:31:18
    • Location: USA Affiliate E5L3CTGE12 Associate 9E88QK5L7811G3H
    • Status: offline
    • Ribbons : 282
    Re: New beta version of core 17 might fix GK104 - can someone test? 2014/04/15 22:02:14 (permalink)
    Testing now looking ok tonight. I thought I would add my 780 to compare.
    I did get 2 P9406 and 1 P9408 on each test set. Both Rigs are i7-3970X OC to 4.3GHz GPU's are NOT OC'ed
    P9406 run nice and complete before 10 Hours
    PM
    Version 0.0.55 Overview Page 
    Status Driver 335.23 Progress Name TPF PPD WUs ETA Core Core ID Project (Run, Clone, Gen) Credit
    NoFrameTimes 0% Foldit 3 x GTX 680 Slot 00 00:00:00 0.00 0.000 00:00:00 ZETA 17 P9408 (R319, C0, G0) 13000
    NoFrameTimes 0% Foldit 3 x GTX 680 Slot 01 00:00:00 0.00 0.000 00:00:00 ZETA 17 P9406 (R211, C0, G0) 10500
    NoFrameTimes 0% Foldit 3 x GTX 680 Slot 02 00:00:00 0.00 0.000 00:00:00 ZETA 17 P9406 (R213, C0, G0) 10500
    Running 1% Foldit 3 x GTX 680 Slot 00 00:08:16 96878.00 1.742 13:38:24 ZETA 17 P9408 (R319, C0, G0) 55615.2
    Running 2% Foldit 3 x GTX 680 Slot 01 00:04:37 105728.10 3.119 07:32:26 ZETA 17 P9406 (R211, C0, G0) 33896.6
    Running 1% Foldit 3 x GTX 680 Slot 02 00:05:21 84624.20 2.692 08:49:39 ZETA 17 P9406 (R213, C0, G0) 31440.3
    Running 4% Foldit 3 x GTX 680 Slot 00 00:08:03 100595.00 1.789 12:52:48 ZETA 17 P9408 (R319, C0, G0) 56235.4
    Running 8% Foldit 3 x GTX 680 Slot 01 00:04:30 109782.00 3.200 06:54:00 ZETA 17 P9406 (R211, C0, G0) 34306.9
    Running 7% Foldit 3 x GTX 680 Slot 02 00:05:00 93834.10 2.880 07:45:00 ZETA 17 P9406 (R213, C0, G0) 32581.3

    *********************** Log Started 2014-04-16T04:02:34Z ***********************
    04:03:56:WU00:FS00:0x17:WARNING:Console control signal 1 on PID 3248
    04:07:01:WARNING:WU00:FS00:FahCore returned: BAD_WORK_UNIT (114 = 0x72)
    04:08:36:WU02:FS00:0x17:WARNING:Console control signal 1 on PID 2128
    04:08:36:WU01:FS01:0x17:WARNING:Console control signal 1 on PID 2624
    04:08:36:WU00:FS02:0x17:WARNING:Console control signal 1 on PID 2216
    04:08:42:WU02:FS00:0x17:ERROR:103: Lost client lifeline
    04:09:06:WU01:FS01:0x17:ERROR:103: Lost client lifeline
    04:09:36:WARNING:FS02:Killing WU00
    ******************************* Date: 2014-04-16 *******************************
    11:51:35:WU03:FS01:0x17:ERROR:exception: Potential energy error of 70.2513, threshold of 10
    11:51:35:WU03:FS01:0x17:ERROR:Reference Potential Energy: -757984 | Given Potential Energy: -758054
    11:51:36:WARNING:WU03:FS01:FahCore returned: BAD_WORK_UNIT (114 = 0x72)
     
     
    AM
    Version 0.0.55 Overview Page 
    Status Driver 335.23 Progress Name TPF PPD WUs ETA Core Core ID Project (Run, Clone, Gen) Credit
    Running 4% Foldit 3 x GTX 680 Slot 00 00:08:03 100595.00 1.789 12:52:48 ZETA 17 P9408 (R319, C0, G0) 56235.4
    Running 8% Foldit 3 x GTX 680 Slot 01 00:04:30 109782.00 3.200 06:54:00 ZETA 17 P9406 (R211, C0, G0) 34306.9
    Running 7% Foldit 3 x GTX 680 Slot 02 00:05:00 93834.10 2.880 07:45:00 ZETA 17 P9406 (R213, C0, G0) 32581.3
    Running 61% Foldit 3 x GTX 680 Slot 00 00:08:16 96775.5 1.742 05:22:24 ZETA 17 P9408 (R319, C0, G0) 55556.3
    Running 4% Foldit 3 x GTX 680 Slot 01 00:08:51 87183.8 1.627 14:09:36 ZETA 17 P9408 (R420, C0, G0) 53581.7
    Running 97% Foldit 3 x GTX 680 Slot 02 00:05:12 88403.4 2.769 00:15:36 ZETA 17 P9406 (R213, C0, G0) 31923.5
     
     
    PM 
    Version 0.0.55 Overview Page 
    Status Driver 335.23 Progress Name TPF PPD WUs ETA Core Core ID Project (Run, Clone, Gen) Credit
    NoFrameTimes 0% Foldit 3 x GTX 780 Slot 00 00:00:00 0.00 0.000 00:00:00 ZETA 17 P9406 (R23, C0, G9) 10500
    NoFrameTimes 0% Foldit 3 x GTX 780 Slot 01 00:00:00 0.00 0.000 00:00:00 ZETA 17 P9408 (R331, C0, G0) 13000
    NoFrameTimes 0% Foldit 3 x GTX 780 Slot 02 00:00:00 0.00 0.000 00:00:00 ZETA 17 P9406 (R150, C0, G7) 10500
    Running 1% Foldit 3 x GTX 780 Slot 00 00:04:03 128675.70 3.556 06:40:57 ZETA 17 P9406 (R23, C0, G9) 36190
    Running 1% Foldit 3 x GTX 780 Slot 01 00:06:40 134053.90 2.160 11:00:00 ZETA 17 P9408 (R331, C0, G0) 62062
    Running 1% Foldit 3 x GTX 780 Slot 02 00:03:28 162216.10 4.154 05:43:12 ZETA 17 P9406 (R150, C0, G7) 39052
    Running 3% Foldit 3 x GTX 780 Slot 00 00:03:53 137096.60 3.708 06:16:41 ZETA 17 P9406 (R23, C0, G9) 36971.7
    Running 2% Foldit 3 x GTX 780 Slot 01 00:06:40 134005.30 2.160 10:53:20 ZETA 17 P9408 (R331, C0, G0) 62039.5
    Running 4% Foldit 3 x GTX 780 Slot 02 00:03:20 172695.00 4.320 05:20:00 ZETA 17 P9406 (R150, C0, G7) 39975.7

    AM
    Version 0.0.55 Overview Page 
    Status Driver 335.23 Progress Name TPF PPD WUs ETA Core Core ID Project (Run, Clone, Gen) Credit
    Running 3% Foldit 3 x GTX 780 Slot 00 00:03:53 137096.60 3.708 06:16:41 ZETA 17 P9406 (R23, C0, G9) 36971.7
    Running 2% Foldit 3 x GTX 780 Slot 01 00:06:40 134005.30 2.160 10:53:20 ZETA 17 P9408 (R331, C0, G0) 62039.5
    Running 4% Foldit 3 x GTX 780 Slot 02 00:03:20 172695.00 4.320 05:20:00 ZETA 17 P9406 (R150, C0, G7) 39975.7
    Running 29% Foldit 3 x GTX 780 Slot 00 00:03:24 167667.9 4.235 04:01:24 ZETA 17 P9406 (R175, C0, G2) 39588.3
    Running 72% Foldit 3 x GTX 780 Slot 01 00:06:40 134037.6 2.16 03:06:40 ZETA 17 P9408 (R331, C0, G0) 62054.5
    Running 40% Foldit 3 x GTX 780 Slot 02 00:03:35 154599.1 4.019 03:35:00 ZETA 17 P9406 (R208, C0, G1) 38470.8

     
    AM
    P9408 Run not so well on my GTX 660Ti
    Version 0.0.55 Overview Page 
    Status Driver 335.23 Progress Name TPF PPD WUs ETA Core Core ID Project (Run, Clone, Gen) Credit
    RunningNoFrameTimes 0% Foldit GTX 660Ti Slot 00 00:00:00 0 0 00:00:00 ZETA 17 P9408 (R435, C0, G0) 13000
    Running 2% Foldit GTX 660Ti Slot 00 00:10:53 64161.2 1.323 17:46:34 ZETA 17 P9408 (R435, C0, G0) 48492.2

     
    PM
    Summary Overview Page 
    Status Driver 335.23 Progress Name TPF PPD WUs ETA Core Core ID Project (Run, Clone, Gen) Credit
    RunningNoFrameTimes 0% Foldit GTX 660Ti Slot 00 00:00:00 0 0 00:00:00 ZETA 17 P9408 (R435, C0, G0) 13000
    Running 2% Foldit GTX 660Ti Slot 00 00:10:53 64161.2 1.323 17:46:34 ZETA 17 P9408 (R435, C0, G0) 48492.2
    Finishing 64% Foldit GTX 660Ti Slot 00 00:10:50 64469.6 1.329 06:30:00 ZETA 17 P9408 (R435, C0, G0) 48501.4

    post edited by bcavnaugh - 2014/04/16 18:22:24

    Associate Code: 9E88QK5L7811G3H


     
    #13
    bison88
    Superclocked Member
    • Total Posts : 136
    • Reward points : 0
    • Joined: 2013/10/28 17:44:21
    • Status: offline
    • Ribbons : 0
    Re: New beta version of core 17 might fix GK104 - can someone test? 2014/04/15 22:51:18 (permalink)
    Just for reference my OC'ed GTX 780 @ 1188Mhz (FAH 24/7 w/Boost) is doing 6 min 01 sec on the 9408.  So the numbers are looking about right depending on the exact clocks w/boost factored in.
    #14
    boballee
    FTW Member
    • Total Posts : 1455
    • Reward points : 0
    • Joined: 2011/03/16 12:19:13
    • Location: Waiting for the Levee to Break at Bron-Yr-Aur
    • Status: offline
    • Ribbons : 8
    Re: New beta version of core 17 might fix GK104 - can someone test? 2014/04/16 05:21:48 (permalink)
    bison88
    Just for reference my OC'ed GTX 780 @ 1188Mhz (FAH 24/7 w/Boost) is doing 6 min 01 sec on the 9408.  So the numbers are looking about right depending on the exact clocks w/boost factored in.




    Good to know.  I'm running a 9408 right now with a TPF of 9:35 - so probably about right on my 660 Ti.  



    #15
    bcavnaugh
    The Crunchinator
    • Total Posts : 38977
    • Reward points : 0
    • Joined: 2012/09/18 17:31:18
    • Location: USA Affiliate E5L3CTGE12 Associate 9E88QK5L7811G3H
    • Status: offline
    • Ribbons : 282
    Re: New beta version of core 17 might fix GK104 - can someone test? 2014/04/16 06:02:26 (permalink)
    bison88
    Just for reference my OC'ed GTX 780 @ 1188Mhz (FAH 24/7 w/Boost) is doing 6 min 01 sec on the 9408.  So the numbers are looking about right depending on the exact clocks w/boost factored in.


    I am not OC my GTX 780 and looking  at 03:06:40 TPF for P9408
    Running 72%  Foldit 3 x GTX 780 Slot 01  00:06:40  134037.6  2.16  03:06:40  ZETA 17 P9408 (R331, C0, G0) 62054.5
     
    Is anyone Running K-BOOT while Folding?
    Or would this reduce the life of the GPU?
     
    I also note that on both the GTX 780 and GTX 680 my wattage usage dropped about 60 Watts Overall on both Rigs.
    post edited by bcavnaugh - 2014/04/16 06:21:00

    Associate Code: 9E88QK5L7811G3H


     
    #16
    yodap
    CLASSIFIED Member
    • Total Posts : 4641
    • Reward points : 0
    • Joined: 2011/05/15 06:13:40
    • Location: NY, Upstate
    • Status: offline
    • Ribbons : 8
    Re: New beta version of core 17 might fix GK104 - can someone test? 2014/04/16 09:44:37 (permalink)
    I installed new beta drivers, deleted old core, reset gpu slot and according to the log, promptly received version 0.0.49. ??? 
    No patience for this.
     
    Edit: Fixed it. I'll post back when I see some results.
    post edited by yodap - 2014/04/16 10:40:46


     

     
    #17
    yodap
    CLASSIFIED Member
    • Total Posts : 4641
    • Reward points : 0
    • Joined: 2011/05/15 06:13:40
    • Location: NY, Upstate
    • Status: offline
    • Ribbons : 8
    Re: New beta version of core 17 might fix GK104 - can someone test? 2014/04/16 11:31:15 (permalink)
    Ok, things are looking good 20% in. 4:03 tpf on a 660ti. The first 6% were before the fix so my 70k estimate is probably a little low. Glad for Pande Group to get this working.


     

     
    #18
    robbysites
    FTW Member
    • Total Posts : 1978
    • Reward points : 0
    • Joined: 2009/01/12 10:29:41
    • Location: 51°10?43.84?N 1°49?34.28?W
    • Status: offline
    • Ribbons : 2
    Re: New beta version of core 17 might fix GK104 - can someone test? 2014/04/16 13:55:28 (permalink)
    I have a GTX 660 folding with the older driver - I take it there would be a benefit to upgrade the driver?


          MY AFFILIATE CODE-000H94333W


     
    #19
    Zagen30
    SSC Member
    • Total Posts : 528
    • Reward points : 0
    • Joined: 2009/11/27 12:21:01
    • Location: Laguna Niguel, CA
    • Status: offline
    • Ribbons : 3
    Re: New beta version of core 17 might fix GK104 - can someone test? 2014/04/16 14:19:18 (permalink)
    I don't think there's much folding benefit to the newer drivers; at best, you might shave a few seconds off the TPF.  The main issue has been those who want or need newer drivers for various games, but were rightfully unhappy with the massive folding hit.


     
    Core i7 3770K @4.3 GHz, 8 GB DDR3 @2133 MHz, 2xGTX 970 SSC @1379 MHz, Windows 7 Pro 64-bit    
     
    Affiliate code: K4BJ4FHUX5
     
    #20
    wrinvert
    R.I.P Friend
    • Total Posts : 3751
    • Reward points : 0
    • Joined: 2007/05/01 19:32:12
    • Location: lost forever
    • Status: offline
    • Ribbons : 17
    Re: New beta version of core 17 might fix GK104 - can someone test? 2014/04/16 15:25:05 (permalink)
    680 2gb on 337.50 running the new core17 .55 3:10tpf is about 105ppd, looks good to me so far.
     
    22:11:58:WU00:FS01:0x17:Version 0.0.55
    22:12:02:WU02:FS01:Upload 71.31%
    22:12:05:WU02:FS01:Upload complete
    22:12:05:WU02:FS01:Server responded WORK_ACK (400)
    22:12:05:WU02:FS01:Final credit estimate, 14093.00 points
    22:12:05:WU02:FS01:Cleaning up
    22:13:06:WU00:FS01:0x17:Completed 0 out of 2500000 steps (0%)
    22:13:06:WU00:FS01:0x17:Temperature control disabled. Requirements: single Nvidia GPU, tmax must be < 110 and twait >= 900
    22:16:18:WU00:FS01:0x17:Completed 25000 out of 2500000 steps (1%)
    22:19:21:WU00:FS01:0x17:Completed 50000 out of 2500000 steps (2%)
    22:22:36:WU00:FS01:0x17:Completed 75000 out of 2500000 steps (3%)

     


     
    #21
    robbysites
    FTW Member
    • Total Posts : 1978
    • Reward points : 0
    • Joined: 2009/01/12 10:29:41
    • Location: 51°10?43.84?N 1°49?34.28?W
    • Status: offline
    • Ribbons : 2
    Re: New beta version of core 17 might fix GK104 - can someone test? 2014/04/16 15:53:44 (permalink)
    "I don't think there's much folding benefit to the newer drivers; at best, you might shave a few seconds off the TPF.  The main issue has been those who want or need newer drivers for various games, but were rightfully unhappy with the massive folding hit."
     
    I knew there was an issue with using the latest driver and if you were folding not being able to use it to game on GTX 770's or lower. I have a 660 and the only reference in the test was a 660ti. I am using 326.80 for the vanilla 660. If you think a newer driver would only shave a second or 2 then it is not worth the trouble. It puts out 50 to 70 k per day. It's slow but it's folding. I have another on the shelf gathering dust.
    On the other hand I have the GTX 780 SC also folding 24/7 using version 334.89 - Would there be any advantage (for folding) to go to a newer driver, and if so what version? Thanks.
    post edited by robbysites - 2014/04/16 15:55:50


          MY AFFILIATE CODE-000H94333W


     
    #22
    wrinvert
    R.I.P Friend
    • Total Posts : 3751
    • Reward points : 0
    • Joined: 2007/05/01 19:32:12
    • Location: lost forever
    • Status: offline
    • Ribbons : 17
    Re: New beta version of core 17 might fix GK104 - can someone test? 2014/04/16 16:21:01 (permalink)
    I updated my 780's both core and drivers today and getting about the same numbers.


     
    #23
    zophar
    CLASSIFIED ULTRA Member
    • Total Posts : 5183
    • Reward points : 0
    • Joined: 2008/08/26 09:48:56
    • Location: KC,MO
    • Status: offline
    • Ribbons : 26
    Re: New beta version of core 17 might fix GK104 - can someone test? 2014/04/16 16:51:46 (permalink)
    Base EVGA GTX760 (1071MHz boost clock, 3005MHz mem clock), running the 335.23 drivers.
    WU: 9101
    Base credit: 4000
    Estimated credit: 19675
    Estimated PPD: 63787
    Estimated TPF: 4min 26sec.
    post edited by zophar - 2014/04/16 16:57:35

    Attached Image(s)



    #24
    bcavnaugh
    The Crunchinator
    • Total Posts : 38977
    • Reward points : 0
    • Joined: 2012/09/18 17:31:18
    • Location: USA Affiliate E5L3CTGE12 Associate 9E88QK5L7811G3H
    • Status: offline
    • Ribbons : 282
    Re: New beta version of core 17 might fix GK104 - can someone test? 2014/04/16 18:25:03 (permalink)
    Zagen30
    I don't think there's much folding benefit to the newer drivers; at best, you might shave a few seconds off the TPF.  The main issue has been those who want or need newer drivers for various games, but were rightfully unhappy with the massive folding hit.


    Then would you recommend going back to the older driver on rigs that are used only for Folding for GK107 cards and below?
    Thanks
    post edited by bcavnaugh - 2014/04/16 18:28:00

    Associate Code: 9E88QK5L7811G3H


     
    #25
    mGuyA
    New Member
    • Total Posts : 45
    • Reward points : 0
    • Joined: 2006/02/04 20:51:52
    • Status: offline
    • Ribbons : 0
    Re: New beta version of core 17 might fix GK104 - can someone test? 2014/04/16 18:36:07 (permalink)
    335.23 driver with core 17, 0.0.55, looking good.
    GK104 GTX 660 TI 915 MHz
    WU: 9101
    Base credit: 4000
    Estimated credit: 20125
    Estimated PPD: 69833
    Estimated TPF: 4 mins 09 secs
    GK104 GTX 660 TI 993 MHz
    WU: 9101
    Base credit: 4000
    Estimated credit: 20401
    Estimated PPD: 73907
    Estimated TPF: 3 mins 58 secs


     EVGA Affiliate Code CR6GOPQR5H
    #26
    Zagen30
    SSC Member
    • Total Posts : 528
    • Reward points : 0
    • Joined: 2009/11/27 12:21:01
    • Location: Laguna Niguel, CA
    • Status: offline
    • Ribbons : 3
    Re: New beta version of core 17 might fix GK104 - can someone test? 2014/04/16 23:36:48 (permalink)
    robbysites
    On the other hand I have the GTX 780 SC also folding 24/7 using version 334.89 - Would there be any advantage (for folding) to go to a newer driver, and if so what version? Thanks.

    That second or two I've been mentioning seemed to come when I upgraded to 337.50 from 335.23, and that was on a pair of 780s.
     
    bcavnaugh
    Then would you recommend going back to the older driver on rigs that are used only for Folding for GK107 cards and below?
    Thanks

    No, I wouldn't.  You'd gain nothing from that.


     
    Core i7 3770K @4.3 GHz, 8 GB DDR3 @2133 MHz, 2xGTX 970 SSC @1379 MHz, Windows 7 Pro 64-bit    
     
    Affiliate code: K4BJ4FHUX5
     
    #27
    RoxPPD
    New Member
    • Total Posts : 12
    • Reward points : 0
    • Joined: 2013/07/27 13:23:44
    • Status: offline
    • Ribbons : 0
    Re: New beta version of core 17 might fix GK104 - can someone test? 2014/04/17 15:12:22 (permalink)
    I concur - 24hrs on my 3x760/Win8 and 1x660Ti/Win7 shows about the same PPD as before. Thanks for all the reseach, guys!

    *I7-3770K @ 3,9Ghz, ASUS P8Z77-V LX, Corsair H110, 2x4 Kingston HyperX DDR3 RAM, GTX 780 + GTX660Ti, Corsair AX 760W
    *I7-920 @ 4,0Ghz, ASUS P6T, Corsair H110, 2x4 Corsair Vengeance DDR3 RAM, HD7970+R9 280X, Corsair AX 860W
    *A4-5300 @ 3,4Ghz, Gigabyte GA-F2A85X-UP4, 2x4 Cruzer Ballistix Sport DDR3 RAM, 2x R9 280X, Seasonic P-1000W
    #28
    madcattt
    Superclocked Member
    • Total Posts : 104
    • Reward points : 0
    • Joined: 2011/03/31 18:00:30
    • Status: offline
    • Ribbons : 0
    Re: New beta version of core 17 might fix GK104 - can someone test? 2014/04/17 18:35:22 (permalink)
    Based upon my numbers I would say the new update works.  Running Core 15's 762x:
    GTX 590s: 7:30-7:45 tpf 24-27k ppd
    GTX 770: 4:20 tpf estimating 229k ppd!


    #29
    the_Scarlet_one
    formerly Scarlet-tech
    • Total Posts : 24581
    • Reward points : 0
    • Joined: 2013/11/13 02:48:57
    • Location: East Coast
    • Status: offline
    • Ribbons : 79
    Re: New beta version of core 17 might fix GK104 - can someone test? 2014/04/18 00:59:02 (permalink)
    GPU:0:GK104 [GeForce GTX 760]
    Base Credit 17123
    Estimated Credit 42512
    Estimated PPD 26700
    Old Driver326 Estimated TPF 24min 30sec New Drivers and Updated core Estimated TPF 22 mins 5 secs 
    Project 13000
    FahCore 0x17
    ETA 1.16 days
     
    I didn't have beta selected when it picked out this little fella, but I guess it actually helped a LOT.  I do not like these 13000 units, as they take days at time.  
    #30
    Page: 12 > Showing page 1 of 2
    Jump to:
  • Back to Mobile