New computer build: Artifacts on screen. (Solved with GPU BIOS update)

[Ryadge]

Edit: Again, Link updated below with 2nd bios bios update that changes the fan curve for the GPU because apperantly, some EVGA GPU's with ACX 3.0 coolers were reportedly... lighting on fire?

http://www.pcgamer.com/evga-to-release-bios-update-to-combat-gtx-10-series-overheating-issues/

Edit: Problem solved. The GPU just needed a bios update and now everything is good.

The problem was widespread with the micron memory that was being used on various GTX1070 GPU's and overclocked cards. Mine was factory overclocked and was getting artifacts only in Microsoft Edge but not Google Chrome. But while gaming, sometimes the PC would hard lockup and i would have to force shutdown the PC.

The GPU was tested in another PC with the same artifacting problem reproduced so the GPU was sent to EVGA for RMA. After reciving a replacement GPU, the same problems were still there.

And then on October 17th, a bios update was posted on EVGA's forums that 'Improved memory overclocking experience for some users' Since my GPU is factory overclocked, i went ahead and applied this vbios update and well, it's been at least a week now and no more artifacting while web browsing and no freezing/hard lockups while gaming anymore.

If anyone else has a factory overclocked EVGA GTX 1070 and is getting these problems, here's a link to the GPU bios update.
http://forums.evga.com/EVGA-GeForce-GTX-10801070-PWM-Operating-Temperature-Update-m2573491.aspx

[Beardface31]

Check for newer drivers, or try a later version. Have you been monitoring your temperatures?

This could be caused by a bad GPU, RAM, or overheating. Try running Windows Memory Diagnostic or memtest86 to check your RAM.

[Bad 💀 Motha]

372.70 is the latest to date (non Beta)
So yea check the other hardware too, such as RAM.
Ensure RAM is setup correctly in BIOS, etc.
Try un-ticking "Hardware Acceleration" in each Web Browser you use; such as IE, Chrome...

[_I_]

post a cpuz validation link
http://www.cpuid.com/softwares/cpu-z.html
cpuz -> validate button -> submit button
it will open a browser, copy the url (address) and paste it here

[Ryadge]

Okay so it should be worth mentioning on Wednesday when playing BF1 beta on this new PC build, my entire PC froze up with loud buzzing noise coming from the speakers (Sound loop) I had to hard reset the PC since Alt+Tab nor Ctrl-Alt-Delete would respond.

When i rebooted the PC, i immediately checked Event Viewer and found there was a error logged in there, mentioning something about Bug_check 0x124 WHEA_UNCORRECTABLE_ERROR. So maybe it is a RAM issue.

This reminded me that for some reason, MRC Fast boot is enabled by default in the UEFI/BIOS settings which i hear can cause stability issues according to this post from another website.

Source: http://www.overclock.net/t/1573289/asus-z170-deluxe-boot-time/60

This is memory training. This option should only be enabled when adamant the memory is stable. Since day one I and others have been trying to explain to the people complaining DDR4 training routines are what cause these longer POST cycles. They are important, oddly not nearly as many people complained about this on x99. Go figure. With MRC fast boot enabled the system will not retrain the memory which can lead to undiagnosed system instability.

Memory should be tested with the proper means first, either HCI Memtest pro or Google Stress App test via Linux Mint.

With this in mind, i went ahead and disabled MRC Fast boot in the UEFI settings and now my PC went from booting up in 12.50 seconds to 19 seconds. (the time it takes to reach the Windows Desktop after pushing the power button) Is an extra 6 seconds gonna bother me? No, not if this makes my PC more stable.

Anyways, i'm still gonna go ahead and run Memtest86 over night to check the ram. I heard 8 passes is good?

[Ryadge]

Oh yeah i forgot to post Tempature reports. Using NZXT's Cam software, Average idle temps for the CPU: 22C. GPU: 41C

Average temps when playing a game (Star Wars: Battlefront in this case) are, CPU: 50-55C. GPU: 72C.

Maximum temp ever reached by the CPU when running the Prime95 blend test was 75C. Max temp i've seen the GPU reach when playing a game was 75C.

Anyways i'm now gonna go ahead and test the RAM using Memtest86 running 8 passes and then i'll report back.

[Bad 💀 Motha]

What Power Supply do you have?
Given what you stated in Post#4, that is usually when happens when a poor PSU is used.

[Ryadge]

Автор сообщения: Bad-Motha

What Power Supply do you have?
Given what you stated in Post#4, that is usually when happens when a poor PSU is used.

EVGA Supernova G2 650w.

[Ryadge]

Edit: After 13 hours 5 minutes, Memtest86 has now concluded after 8 passes with no errors. Note i did disable MRC Fastboot in the UEFI before running this so i may want to re-enable MRC Fastboot and then re-run Memtest86 again to see if this was the problem.

[Bad 💀 Motha]

MRC Fastboot and the FastBoot within Win10 should just be disabled, regardless.

[Ryadge]

Bumping this old thread to say the problem has been solved with a GPU BIOS update from EVGA. See my edited post at the top of this page for details and a link if you have this problem too with the same GPU. Even if you don't, you may want to update it anyways if yours uses EVGA's ACX 3.0 cooler since apperantly, some EVGA GPU's with ACX 3.0 coolers are reportedly lighting on fire under extreme stress. http://www.pcgamer.com/evga-to-release-bios-update-to-combat-gtx-10-series-overheating-issues/

If you're using the Founders edition GTX1080 or 1070, you do not need this update.