Black Screen GTX 970 (Getting Frustrated)

Dranej_1

Commendable
May 17, 2016
4
0
1,510
I usually get black screen while playing games unless if i don't increase the fan speed with 50 percent and 75 percent if i play gta v. Today, I tried not to increase the fan speed since i will only watch youtube videos and browse and guess what...

What i've seen todays issue:
* 1. Straight Black screen and needs to reset
* 2. Black Screen Flashes and Driver not responding (after crashed i checked oph and after burner http://imgur.com/a/YJ0uV)
* 3. Screen turns yellow and artifacts shows bottom of the screen and needs to reset.

What i've done so far with MSI After Burner
* Increase Fan speed by 50-75 (Problem still persist sometimes on very high intensive games such as gta v )
* Increase Core Voltage by 3-6 (Didn't help at all)

I'm still planning to rma the card or troubleshoot futhermore.

EDIT: Only happens if gpu temp is 65c or goes beyond that.

Specs:
Asus Strix GTX 970 Factory OC'ed (Using latest driver and bios version)
Asus B85 Pro Gamer
i7 4790K (NOT OC'ed)
h100i gtx
SeaSonic X Series X650w Gold
1 TB HDD
16 GB Ram
250GB SDD
NZXT H440

 
Solution
I would keep the card at those clocks while gaming for a few hours and if that fixed it i'd tell the manufacturer about the issue and claim your warranty since its most likely because of your chip.


7 months?, it was built on december 26 2016 and I also got black screen on first play with World of warcraft, then i began to set the custom fan speed.
 
I had to make a custom BIOS for my Gigabyte G1 GTX 970 to stop the driver crashes in the midrange boost states. NVidia's default voltage tables are too aggressive toward power savings. On the mid ASIC cards binned for overclocking, this can cause instability.

You could see if there is a custom BIOS for your card that bumps the stock voltage table up by .0125-.025 per step. this fixed mine perfectly. Hardware wise, it's been rock solid stable for almost 2 years now. And tweaking the BIOS even allowed me to build in a custom fan curve that keeps it below 65c @ 1506 core under full load and is still QUIET.

If you want to read more about what the community is doing about this, here are some links:

Mine:
http://www.overclock.net/t/1556760/mad-dog-fargos-gigabyte-g1-gaming-bios-hynix-ram-final

ASUS Strix by Zoson:
http://www.overclock.net/t/1517316/extract-and-flash-gtx-970-and-980-firmware-zosons-h2o-and-air-custom-bios/2460


*** BE AWARE!!! When done right, this works great. If you have a dual BIOS card, it is easier to recover from a bad flash. BIOS can vary depending on your card's RAM. You can also extract your own stock BIOS and make minimal tweaks to is yourself. It's worth learning how to do. :)
 


Its great you solved it, but to be honest, its unacceptable for these card makers to assume everyone will make their own custom bios to fix their issues. Its like selling a motherboard and leaving out random components leaving it up to the consumer to solder it.

If your comfortable doing this, go for it. If not, don't bother as it may void your warranty and you can be left with a bricked card.
 
UPDATE: I uninstalled nvidia latest driver and custom installed 353.49 with perform clean install, I manage to get no black screen for about an hour playing gta v, the temperatures were fine the card didn't go above 71c. I shut down the computer and let it rest for about 10 minutes, then i turn it back same thing again black screen occurs and restarted itself. I shut it down again for couple of minutes, played CSGO no black screen then i tried GTA V for 2 hours without a problem. But still problem persist if don't set custom fan speed, It still also occurs while browsing and watching youtube.
 
I think your gpu is unstable. Thats usally what happens to me if i push the clock to its limit. The clock is getting more and more unstable on higher temps but that really shouldnt happen on stock clocks. Atleast not under 100°C 😀

Maybe try to underclock your card a little and see if it gets stable and check the vram with this tool. Also try to put the card into another computer and benchmark it with heaven benchmark and/or furmark.
If underclock solves it or your mem is faulty i would claim warranty since you got a bad chip. If it works fine in another computer (where it doesnt get bottlenecked too much by cpu) its most likely your systems fault. Then reinstall windows, do a memtest, check cpu stability etc...

Thats what i would do now.
 
I underclocked my card by -64 core clock and -65 memory clock, It run Heaven benchamark without black screen.

With Video Memory Test Log:
Changing video mode to 640x480x16...ERROR (Code: -2)
[5/18/2016 10:05:50 PM] Test started for "Primary Display Driver (NVIDIA GeForce GTX 970)"...
Trying 32bpp RGB:888 mode...OK
Trying 32bpp BGR:888 mode...NOT SUPPORTED (Code: 80004001)
[5/18/2016 10:07:49 PM] Pass completed (0 errors found).
 
I would keep the card at those clocks while gaming for a few hours and if that fixed it i'd tell the manufacturer about the issue and claim your warranty since its most likely because of your chip.
 
Solution