cancel
Showing results for 
Search instead for 
Did you mean: 

ASUS Turbo RTX 2080Ti dead or something else?

Sirrrus
Level 7
Hello, here’s my rig I just bought 2 weeks ago:
*
Mobo: MSI MEG Z390 ACE
Processor: INTEL CORE i9 9900k 8C/16T 5.0GHz OC
RAM: DDR4 3200MHz G.SKILL Trident Z (2x16GB)
PSU: Corsair RM1000X Gold Rated
GPU: ASUS RTX 2080Ti
SSD: Samsung 970 PRO NVME 1TB
Heatsink: CORSAIR Hydro Series H100i PRO
OS: Windows 10 Professional 64-Bit
Case: Lian Li PC-011 WX
UPS: CyberPower 1500PFCLCD
Monitor: SPECTRE C32 (32” @ 144Hz and 1080p)
Been have been having issues right out of the box. This rig is solely for gaming and occasional web browsing. All drivers were fully up to date BEFORE running any games. The game causing the most consistent issues is Witcher 3, although other games have produced the same problems (AC Odyssey, Battlefield V, Fallout 4, etc.) none as consistently as Witcher 3. I can play for about 20 mins before I get screens with artificing and odd color blotches of black, fuchsia, lime green, etc. I also hear the sound stop and stutter until the PC must be shut down. If I wait long enough, the screen detects “No Input”. This led me to believe it was the video card. I had GPU Tweak II installed with a custom fan curve to ensure proper cooling but monitoring the temps showed nothing even close to maxing on any part of the hardware, all were at least 15° C away or more from max under load. However, after shutting down during these graphics issues, I was informed upon reboot that Windows needed Repaired, which got me thinking corrupted Windows install. It was unable to repair so I had to wipe the hard drive and USB Windows 10 install again. After reloading Windows and re-installing games, I tried again with the same results. I have also tried re-seating the GPU and RAM. Does this sound like a defective video card or am I looking at another issue, or worse, several issues?

Thank you in advance for your help, feel like I’ve lost my child. **
4,349 Views
14 REPLIES 14

JustinThyme
Level 13
Its that Cyberpower UPS! J/K...Im an engineer for Schneider Electric in the critical power division. AKA APC and MGE UPS systems.
Is it overclocked at all? Artifacts usually start as a result of poor memory performance. Just because its not hitting thermal limit on the core doesnt mean the memory isnt failing after a bit of time from heat. Hows your case air flow?

The windows thing is probably a result of the hardware fault and can be alleviated by disabling fast start up. Problem is if you have a problem at shutdown it just reloads the same problem and there is nothing fast about it. Not a bit of difference other than causing headaches.

My suggestion is to try just uninstalling GPU tweak, wipe it with Ccleaner so there are no remnants left and see how it acts at defaults. Make sure you have sufficient air intake to the case too. These are hot cards, almost too hot to run on air IMO but if you do you need good airflow coming in to the GPU.

Also once you get GPU tweak off the chart a fresh install of the latest drivers wont do any harm.



“Two things are infinite: the universe and human stupidity, I'm not sure about the former” ~ Albert Einstein

Sirrrus
Level 7
Thanks for quick reply. The Proc is Overclocked but the card is not. I was going to run it stock and see how things went before pushing it. I have monitored the temps and they never really got over 60° C in Witcher 3, but got higher in other games. The card safe limit I believe is 84° C, so it shouldn’t have a problem. I did notice after just getting 1 successful reboot without a Windows re-install that Device Manager cited a problem with the GPU and shut it down. So I’m pretty sure it’s the GPU but shouldn’t I be able to load Windows Ok on the next reboot if it disables the GPU? Headed back home now, I’ll try the GPU tweak removal however I custom set the fans to run at slightly higher RPM’s than the stock settings. If overheating is the issue then it will likely do so sooner. Is it possible to overheat around 60° C and crash when it would not do so at higher temps?

Silent_Scone
Super Moderator
1. Reseat the GPU and associated power cables

2. Make sure the card is operating at default frequency

3. With no manual overclock applied there should be no instability below the temperature target. (84C).
13900KS / 8000 CAS36 / ROG APEX Z790 / ROG TUF RTX 4090

JustinThyme
Level 13
I don’t think it’s the GPU as much as the VRAM. In my experience the GPU gets hot or hits power limit it just throttles. If an OC is unstable it will crash to desktop or end whateveryou are doing. If you get artifacts and freezing that’s pretty much a safe bet it’s VRAM. I’m not overly trusting of the latest GPU tweak. Very buggy. Had my Strix 2080Tis Nvlink doing strange things. I had similar things happen, not the windows issue, and uninstalled, went through a few reboots doing other things like trying Precision X1 amd After *burner and uninstalling and wiping those because it looks like they can’t communicate with the Strix cards. I could move power target and core but not voltage*and the settings would not apply. Everything else grayed out. Unfortunately ASUS still sells cards labeled OC that are not OC’d out of the box like everyone else. You have to use software that has been plagued by bugs since it’s inception to get the “Stock OC” I’d be happy just to run the cards at the speed I bought and paid a premium for with no software needed then maybe just for fun benching for bragging rights use a software tool that’s reliable because as it stands 2X 2080 TI cards with 100GB/S Nvlink will crush anything you throw at it. I’ve been beating*
*The crap out of them while they are still on air before I yank these monsterous heatsinks off and out water blocks on. While benching I’ve had to leave my case open because the heat they dump in the case it unreal. Here I am running benches for GPUs that are not at all CPU intensive yet out of nowhere a loop with 2X 480 rads and a 360 with no load but the CPU that isn’t loaded starts creeping up in fan speeds. The heat from the cards was getting pulled through my rads increasing 2 liters of coolants temps enough to make it rise 4C and climbing. The top card that has hindered air flow even though they are 4 slot spaced runs a full 10C hotter than the bottom. I’ve already had one heatsink off and the VRAM as usual get squat for cooling. In my cards it’s the brace being used. None of the chokes get cooling. Knowing that when that too card was loaded and the back plate so hot I couldn’t touch it I have no doubts the VRAM was too hot. All this in a monster case, Enthoo Elite. Blocks will be in tomorrow and they will be going under water .**



“Two things are infinite: the universe and human stupidity, I'm not sure about the former” ~ Albert Einstein

Silent_Scone
Super Moderator
JustinThyme wrote:
I don’t think it’s the GPU as much as the VRAM. In my experience the GPU gets hot or hits power limit it just throttles. If an OC is unstable it will crash to desktop or end whateveryou are doing. If you get artifacts and freezing that’s pretty much a safe bet it’s VRAM. I’m not overly trusting of the latest GPU tweak. Very buggy. Had my Strix 2080Tis Nvlink doing strange things. I had similar things happen, not the windows issue, and uninstalled, went through a few reboots doing other things like trying Precision X1 amd After *burner and uninstalling and wiping those because it looks like they can’t communicate with the Strix cards. I could move power target and core but not voltage*and the settings would not apply. Everything else grayed out. Unfortunately ASUS still sells cards labeled OC that are not OC’d out of the box like everyone else. You have to use software that has been plagued by bugs since it’s inception to get the “Stock OC” I’d be happy just to run the cards at the speed I bought and paid a premium for with no software needed then maybe just for fun benching for bragging rights use a software tool that’s reliable because as it stands 2X 2080 TI cards with 100GB/S Nvlink will crush anything you throw at it. I’ve been beating*
*The crap out of them while they are still on air before I yank these monsterous heatsinks off and out water blocks on. While benching I’ve had to leave my case open because the heat they dump in the case it unreal. Here I am running benches for GPUs that are not at all CPU intensive yet out of nowhere a loop with 2X 480 rads and a 360 with no load but the CPU that isn’t loaded starts creeping up in fan speeds. The heat from the cards was getting pulled through my rads increasing 2 liters of coolants temps enough to make it rise 4C and climbing. The top card that has hindered air flow even though they are 4 slot spaced runs a full 10C hotter than the bottom. I’ve already had one heatsink off and the VRAM as usual get squat for cooling. In my cards it’s the brace being used. None of the chokes get cooling. Knowing that when that too card was loaded and the back plate so hot I couldn’t touch it I have no doubts the VRAM was too hot. All this in a monster case, Enthoo Elite. Blocks will be in tomorrow and they will be going under water .**


It's best to cover all bases before going off half-cocked. If it is the VRAM then the card will need to be RMA.

Also, the OC profile is applied by default. There is a more aggressive alternative which can be applied manually via GPUT, however, the card is in fact factory overclocked.

I was able to use the 2080 Strix with AfterBurner without issue, so I'm not sure why it wouldn't work for you.
13900KS / 8000 CAS36 / ROG APEX Z790 / ROG TUF RTX 4090

JustinThyme
Level 13
Here we go again...
Did you read my first reply that you repeated almost verbatim with your first post? No one said the VRAM was faulty, just that it could be a related setting as artifacts are nearly always attributed to Vram.

Have you run a strix 2080ti 011G? *The OC is NOT applied by default out of the box and was that way with pascal 1080ti OC version as well and neither was the 980 ti OC version applying the specified OC overclock out of the box This has been a complaint for a VERY long time and it’s still there. No software no Flagship OC profile. You get 1650 which is the underclocked gaming profile. Granted you can get even more with the software but you can’t get what’s on the spec sheet of the Strix 2080Ti 011G Gaming without it.*

I don’t know why the other software doesn’t work either, it did on pascal but not this time. *Maybe it’s platform specific? I don’t know. All I know is two cards with Nvlink do not work woth anything but GPU tweak on an X299 R6E bios 1603 with a 9940x running win 10 pro 1809.


Ill let GPUZ do the talking, please tell me where in the shot without GPU tweak running you see the factory OC of the Strix O11G?

78384

Now again with gpu tweak running and factory OC selected

78385


Now again with the OC Ive achieved thus far on air

78386



“Two things are infinite: the universe and human stupidity, I'm not sure about the former” ~ Albert Einstein

JustinThyme
Level 13
Now back at OP, have you tried with no GPU tweak and just a fresh driver install?
When you say your CPU is OCd does that include and change to the BCLCK? If its not 100MHz often other peripherals dont play nicely even when the PCIE bus is supposed to be locked at 100MhZ



“Two things are infinite: the universe and human stupidity, I'm not sure about the former” ~ Albert Einstein

Silent_Scone
Super Moderator
JustinThyme wrote:
Here we go again...
Did you read my first reply that you repeated almost verbatim with your first post? No one said the VRAM was faulty, just that it could be a related setting as artifacts are nearly always attributed to Vram.

Have you run a strix 2080ti 011G? *The OC is NOT applied by default out of the box and was that way with pascal 1080ti OC version as well and neither was the 980 ti OC version applying the specified OC overclock out of the box This has been a complaint for a VERY long time and it’s still there. No software no Flagship OC profile. You get 1650 which is the underclocked gaming profile. Granted you can get even more with the software but you can’t get what’s on the spec sheet of the Strix 2080Ti 011G Gaming without it.*

I don’t know why the other software doesn’t work either, it did on pascal but not this time. *Maybe it’s platform specific? I don’t know. All I know is two cards with Nvlink do not work woth anything but GPU tweak on an X299 R6E bios 1603 with a 9940x running win 10 pro 1809.


Ill let GPUZ do the talking, please tell me where in the shot without GPU tweak running you see the factory OC of the Strix O11G?

78384

Now again with gpu tweak running and factory OC selected

78385


Now again with the OC Ive achieved thus far on air

78386



Justin, the GPU Tweak shots you have provided show 1650MHz, which is is a factory overclock. Reference FE cards have a factory OC of 1635MHz, as can be found easily on the NVIDIA website: https://www.nvidia.com/en-gb/geforce/graphics-cards/rtx-2080-ti/

Perhaps your understanding of VRAM is different to mine, however, if the memory is causing artifacts at stock operation then I’m not sure how it’s possible to conclude it is not faulty (my post reiterates making sure the card is at default settings). I felt it was needed to reiterate as your post is condensed and isn't easy to read.

If you can attach a dxdiag to your post, I can see if it's possible to replicate the AB issue, however on the X299 Apex I had no problems using the Strix 2080 RTX, so can only assume it's a local issue or on AB side.
13900KS / 8000 CAS36 / ROG APEX Z790 / ROG TUF RTX 4090

Silent Scone@ASUS wrote:
*Perhaps your understanding of VRAM is different to mine, however, if the memory is causing artifacts at stock operation then I’m not sure how it’s possible to conclude it is not faulty (my post reiterates making sure the card is at default settings). I felt it was needed to reiterate as your post is condensed and isn't easy to read


Fair enough on the repost

Now let’s talk about VRAM.
I know at times I am overly verbose but only because I don’t like making statements that are half understood leading to more questions and more posts.

In my statement I shared how I had the same issue with GPU tweak being the culprit. It lost its mind and even resetting to defaults the entire platform became unstable. The fix? Uninstall GPU tweak, wipe it, the. Reboot without it. Problem solved. After trying the others and both not working as expected I reinstalled GPU tweak and other than the usual bugs of profiles disappearing and the measurements not binding to the settings it’s semi working.

Now if after canning all software related to GPU OC and a new driver install I would whole heartedly agree the card would need to be RMA’d *



“Two things are infinite: the universe and human stupidity, I'm not sure about the former” ~ Albert Einstein