EVGA

Quad 980 Ti setup (Not SLI), 4th card constantly running at Boost(?)

Author
Post
Essentials Only Full Version
Sikotik13
New Member
  • Total Posts : 3
  • Reward points : 0
  • Joined: 2008/11/20 13:30:50
  • Status: offline
  • Ribbons : 0
2019/02/03 20:47:42 (permalink)
Hi there,
 
My rig has 4x 980 Ti's, not running in SLI (don't even have the bridge thing connected as I would never use it), intended for rendering, etc.
I'm not really sure when it started, but I recently did a reinstall of Windows out of frustration with this and other issues (I assumed due to driver conflicts from updating/reinstalling so many graphics things over time).
All of the other issues are gone, but the 4th card is (still) always running at GPU Clock 1101, Mem. Clock at 3305 (per Precision, or 1652 according to HWInfo). Can't remember offhand if this is Boost or Base clock speeds, but either way, it ain't idle, as it should be.
All displays are connected to card 1 (or 0, if you're a system view person), and that one goes up and down dynamically as I do stuff, cards 2 and 3 (or 1 and 2) sit at the idle state unless I launch something that specifically utilizes them.
Card 4 (or 3) should exhibit the same behavior as 2 and 3 (1 and 2), and yet it constantly "runs."
There is nothing on the computer to run it independently, I've done every mal/spy/adware sweep I can think of. Nothing anywhere in Windows shows as using that card, and if I attempt to do anything with it (as I do for the middle 2), It's pretty much an instant lock up and (roughly 2 minutes later) crash. 
 
Kinda frustrated at this point. Any advice, ideas are welcome. Thanks in advance.
 
Oh, system specs:
Asus X99-E WS
Intel(R) Core(TM) i7-5960X CPU @ 3.00GHz (16 CPUs), ~3.0GHz (Water-cooled and currently OC'd a bit, but I have troubleshot at stock clocks as well, the issue persists.)
64 GB RAM
4x SSDs
2x HDD
4x EVGA Geforce 980 Ti's (no overclock aside from how they ship, never needed to, also all water-cooled)
Visual of what goes on:
 
post edited by Sikotik13 - 2019/02/03 20:51:07

Attached Image(s)

#1

5 Replies Related Threads
    Sajin
    EVGA Forum Moderator
    • Total Posts : 49168
    • Reward points : 0
    • Joined: 2010/06/07 21:11:51
    • Location: Texas, USA.
    • Status: offline
    • Ribbons : 199
    Re: Quad 980 Ti setup (Not SLI), 4th card constantly running at Boost(?) 2019/02/04 16:15:26 (permalink)
    See if disconnecting all monitors but one will help fix the issue. Yes, I know your monitors aren't connected to that card.


    See my current build here: https://forums.evga.com/My-bench-rig-unlimited-power-m3596024.aspx
    RTX 2080 Ti 4-way SLI?!?!: https://forums.evga.com/RTX-2080-Ti-4way-SLi-set-by-yours-truly-m2954756.aspx
    #2
    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: Quad 980 Ti setup (Not SLI), 4th card constantly running at Boost(?) 2019/02/04 17:34:20 (permalink)
    Looks like that is the GPU that a Monitor is connect too.
     
    Spaces are Pain in Links
    https://forums.evga.com/download.axd?file=0;2917269&filename=GPU Issue.PNG
     
    post edited by bcavnaugh - 2019/02/04 17:37:56

    Associate Code: 9E88QK5L7811G3H


     
    #3
    Sikotik13
    New Member
    • Total Posts : 3
    • Reward points : 0
    • Joined: 2008/11/20 13:30:50
    • Status: offline
    • Ribbons : 0
    Re: Quad 980 Ti setup (Not SLI), 4th card constantly running at Boost(?) 2019/02/04 18:13:08 (permalink)
    @Sajin, the issue seems to have corrected itself, for the moment at least. I didn't do anything, but it is now idling normally. No restart, nothing. Just left it alone and went to bed, and it was fine today as mysteriously as it was messed up. Go fig. If/when it does it again, I will definitely give your suggestion a shot.
     
    @bcavnaugh, sorry about the file name, it's a habit for some of the resources I use to differentiate them. I'll try to keep it in mind when I post images. Also, no. All monitors are connected to the top one (the only one using any VRAM and memory in that shot). No monitor has ever been connected to that card, which was part of why it was confusing that the issue arose. The physical connections etc. on this build haven't changed aside from a few HDD and CPU cooler swaps in about 3 years since I installed the 980s to replace 780s.
     
    I'll keep an eye out for it doing it again, and go from there. Thanks for the input, y'all!
    post edited by Sikotik13 - 2019/02/04 18:15:39
    #4
    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: Quad 980 Ti setup (Not SLI), 4th card constantly running at Boost(?) 2019/02/05 15:03:15 (permalink)
    When I was running 4-Way SLI every time I updated the NVIDIA Drivers it you cause the order of my cards to change.
    Not sure if this helps or not but my Bottom Card became the Primary call the Time as shown in software but still the monitor was connected to the Top Card next to the CPU. 

    Associate Code: 9E88QK5L7811G3H


     
    #5
    Sikotik13
    New Member
    • Total Posts : 3
    • Reward points : 0
    • Joined: 2008/11/20 13:30:50
    • Status: offline
    • Ribbons : 0
    Re: Quad 980 Ti setup (Not SLI), 4th card constantly running at Boost(?) 2019/02/05 16:15:25 (permalink)
    I don't doubt it, I have seen some... odd things after driver updates (and some recurring, which is why I removed the SLI bridge entirely, as it was more aesthetically pleasing, but every update made weird things happen, even when unused). I simply know everything is connected to the top one because at the time I took the image, I had a graphics program open, and everything for that card changed as the displays loaded things (like Chrome, to write the original post). The lower one ran like that constantly, regardless of what I did/didn't do, never actually using any VRAM (above what the other 2 idling cards use, anyway) or memory controller, and if I attempted to utilize it with say, a render, it would crash the computer as soon as it tried to load anything into the VRAM. Not sending info to it through program control (which thankfully let me deselect it so I could get something done), everything would run fine. Still boggles me how it happened, but at least I have a thing to try if it occurs again.
    #6
    Jump to:
  • Back to Mobile