Tech Support Forum banner
Status
Not open for further replies.

GTX 770 Uses 102% Power under load, then crashes

3K views 25 replies 2 participants last post by  greenbrucelee 
#1 ·
Hi everyone,

I hope you can help me. I am running a GTX 770 4GB card. Just recently it has started crashing when playing games.

When I say this, I mean the card runs fine for a short while (2 mins or so) and then all power to the board is lost. The PC stays on, but obviously the screen goes black because there's no signal from the GPU. At this point, the fan speed hits 100%. After a few seconds I just held down the power button for a few seconds until the computer would restart.

I have used GPU-Z to track various statistics, and it seems as though the last entry shows power consumption at 101.8% (obviously it should limit itself to 100, but isn't for some reason).


Have you any idea what is wrong?

This has generally happened when I have been running games with no FPS limit. I am a computer science student writing my own game atm, and that caused it to crash - it didn't have an FPS cap. I was running GRID 2 the other day with no FPS cap - it was running at about 1000FPS on the main menu and crashed. As soon as I invoke the use of DXTORY to limit the FPS, the problem doesn't seem to happen. It does not happen when Idle.

This is fine, but I would really like to stop the card from going over 100%!!

Thanks in advance for your help!


You can (if you choose) download the .txt file log from GPU-Z at the time of the crash. The 101.8% is the last entry:
Download GPU Crash 25.11.14.txt
 
See less See more
1
#5 · (Edited)
Thanks for your reply. My specs are:

  • Intel® Core™i7 Quad Core Processor i7-4770 (3.4GHz) 8MB Cache
  • ASUS® Z87M-PLUS: m-ATX, USB3.0, SATA 6.0, XFIRE
  • 16GB KINGSTON HYPER-X FURY DUAL-DDR3 1600MHz, X.M.P (2 x 8GB)
  • Palit 4GB NVIDIA GEFORCE GTX 770
  • 1TB SEAGATE ST1000DX001 1TB/8GB SSHD 64MB HYBRID DRIVE
  • CORSAIR 750W RM SERIES™ MODULAR 80 PLUS® GOLD
  • Creative Sound Blaster Z 5.1 PCI-E Soundcard

I would say that it more than possible that the card went more than 102%, but it wasn't recorded in the .txt file because the computer had shut down OR communication with the hard drive was terminated before shutting down.


How would I obtain "voltages and temps from your BIOS", please?
 
#6 ·
you go into your BIOS (thats the bit were it says press del or whatever) when you first boot your system and look for the voltages that relate to the 12v, 5v and 3.3v and also your temps. Write it down and post them back or take a photo and upload it.
 
#11 ·
It appears as though (roughly) when the crash happened on 25th, there is an event log entry under "Critical" which reads:

Kernel-Power: The system has rebooted without cleanly shutting down first. This error could be caused if the system stopped responding, crashed, or lost power unexpectedly.
There's another entry under "Warning" which reads:

Kerne-PnP: The driver \Driver\WUDFRd failed to load for the device ACPI\PNP0A0A\2&daba3ff&1.
I say "roughly" because I know it was late afternoon, and it's showing as 16:31, so that would tie in.

Exactly the same entries appear an hour before when it also shut down.
 
#13 ·
What Driver version should it be on?

I have found my "ACPI x64-based PC" under "Computer", and it says the driver is version "6.3.9600.16384" which is apparently from 2006!!!!

When I click the automatic search for updates, it says it's up to date. Where can I find the latest version online?
 
#14 ·
go to the asus website and install all the latest drivers from there and see if that helps.

when you install the drivers from the motherboard disc I always recommend updating to the latest ones from the site either through the asus software or manually.
 
#19 · (Edited)
I just tried it again (with my game, no FPS limit). The Power Consumption went up and up until it hit (in this instance) 90% before the screen went black.

It took about 5-10 seconds after going black before the computer restarted. New entries in the Event Viewer were logged:

Kernel-Power: The system has rebooted without cleanly shutting down first. This error could be caused if the system stopped responding, crashed, or lost power unexpectedly.
No entry was made at the exact time the last GPU-Z log was made.



I would add, the version of the ACPI has not updated since running the driver update. Again, all that happened was a CMD window popped up for 10 sec or so, and that was it.



If you want to download the .txt log file, please feel free. Download GPU Crash 27.11.14.txt

Edit: I've noticed it crashed when GPU Load was at 99% on 27th, and at 97% on 25th. Could this be a contributing factor? Also, VDDC is not getting above 1.2000v ever, and in both documented crashes, they have been at 1.2000v before dropping.
 
#23 ·
since you mentioned the voltage it looks like the card may not be using the power needed to run properly and is possibly overvolting with the 110% etc usually when 110% or more isued that is when someone is overclocking the card.
 
#24 ·
have been speaking to someone at Asus about this as I have ties with them. The guy I spoke to reckons it is power related even though your voltages look ok. He reckons that the psu is spiking when the GPU is underload and will ocassionally do it when its idle.

So I suggest if you can borrow a good psu to try and see what happens.
 
#25 ·
Thank you very much for your helpful advice. Yesterday I phoned up the manufacturer and arranged to send my GPU back.

They'll run some tests - if it comes back with no problems, I shall send off my PSU. I'll update you as it happens. Luckily I have an integrated GPU which I can use to do work with in the meantime.
 
Status
Not open for further replies.
You have insufficient privileges to reply here.
Top