• Ever wanted an RSS feed of all your favorite gaming news sites? Go check out our new Gaming Headlines feed! Read more about it here.
  • We have made minor adjustments to how the search bar works on ResetEra. You can read about the changes here.
Jun 1, 2018
4,523
ComputerBase has provided a quick overview of what users are reporting. At LinusTechTips forums users are reporting problems with ZOTAC GeForce RTX 3080 Trinity. Owners of MSI GeForce RTX 3080 Ventus 3X OC are also noticing this type of behavior, a thread on Overclockers.co.uk seems to suggest. Even on ComputerBase's own forums, there are threads describing the same problem.

This now clearly widespread, but still unexplored issue appears in a certain group of games when the boost clock exceeds 2.0 GHz. As soon as the clock speeds reach a certain level, the game crashes to desktop.

videocardz.com

GeForce RTX 3080 sees increasing reports of crashes in games - VideoCardz.com

An increasing number of users are reporting the crash to desktop (CTD) issues with factory-overclocked NVIDIA GeForce RTX 3080 graphics cards. The issues were reported on various forums and social media platforms. Users are reporting a crash to desktop issues with custom GeForce RTX 3080 models...

NVIDIA still didnt comment on this and the mods on the nvidia subreddit keep deleting posts that mention those crashes.

EDIT:
This has been fixed as of today by the latest nvidia drivers! Yaaay!
 
Last edited:
Apr 4, 2018
4,514
Vancouver, BC
Dang,
I hope Nvidia can reign this in. I was hearing well before these cards came out, that the fragmentation of design in Nvidia cards was causing a lot of complexity (in terms of driver engineering on Nvidia's end), and in turn, less stability. Kind of not surprised here.
 

Theswweet

RPG Site
Verified
Oct 25, 2017
6,418
California
I wonder if it has to do with memory temps/stability? There were reports that nVidia originally wanted to go with even faster chips, but then they had to downclock them... that, and I know at least MSI has had a problem with properly cooling memory components in the past.
 

Sedated

Member
Apr 13, 2018
2,598
Subreddit mods deleting posts? Are they paid by nvidia?

Edit: Saw a thread there on it so mods probably deleting multiple posts on same topic and keeping it to one thread.
 

Hasney

One Winged Slayer
The Fallen
Oct 25, 2017
18,636
Haven't seen anything on my MSI Trio X and it goes above 2Ghz. Looks like dropping by 50Mhz solves it, so I wonder if it's a power delivery issue in the GPU BIOS requesting more power than it can take.

Between this and the paper launch, they really came in hot.

EDIT: looking at what cards this has been reported on, these two and the FE using the in box power adaptor, it does seem to be a power issue since it's only affecting the 2 power connector card. At least that's my theory.
 

Linus815

Member
Oct 29, 2017
19,789
There were several reports saying that AIBs had very little time to prepare the first batch of cards, so it could be a power delivery issue, if not a simple driver problem.
 
Nov 2, 2017
2,275
This now clearly widespread, but still unexplored issue appears in a certain group of games when the boost clock exceeds 2.0 GHz. As soon as the clock speeds reach a certain level, the game crashes to desktop.
Doesn't this just mean the cards are running an unstable overclock? If it does those clocks out of the box then it seems pretty easy to fix where you just limit the boost more.
 
Oct 26, 2017
1,030
I wonder if the cards are pushing it a bit more than usual and the not-so-great cards can't ride the line well. IIRC, Zotac in particular doesn't have a great reputation.

This also lines up with Steve from Gamers Nexus finding that the 3080s don't overclock well.
 

TAJ

Banned
Oct 28, 2017
12,446
I guess not being able to get an order in was a blessing in disguise.
 

soniko_

Banned
Jan 25, 2018
178
People are buying the top of the line video card ... and they want the overclocked one?

jesus effin christ
 

Super Rookie

Member
Oct 25, 2017
276
London
Looks like I lucked out on the Gigabyte Gaming OC card, no issues apart from G-Sync with C9.
Some comments are suggesting they've experienced similar issues with previous partner cards and downclocking fixes it.
 

alphacat

One Winged Slayer
Member
Oct 27, 2017
4,932
*hugs stupid overpriced 2080Ti*

guess even more reason for me to wait for 4X cards
 

ara

Member
Oct 26, 2017
13,020
This is why you wait for Ti and Super folks!!

I agree

The reason is clearly the frankly pitiful amount of VRAM on these cards

So stay away from the 3080s (especially MSI's Gaming X Trio, that particular card is terrible and nobody should buy it) and wait for the ti and/or Super versions, people
 

Deleted member 12317

Account closed at user request
Banned
Oct 27, 2017
2,134
I had a similar problem with my 1080 Ti a few years ago, the boost was too agressive and it resulted in crashes when the core speed went too high by its own.
Had to lower the default core speed by -100MHz, then the boost would still go very high but not high enough to crash.

They should make the boost less agressive.
 

Finaika

Member
Dec 11, 2017
13,330
I agree

The reason is clearly the frankly pitiful amount of VRAM on these cards

So stay away from the 3080s (especially MSI's Gaming X Trio, that particular card is terrible and nobody should buy it) and wait for the ti and/or Super versions, people
What's wrong with the Gaming X Trio?
 

Dries

Banned
Aug 19, 2019
309
I agree

The reason is clearly the frankly pitiful amount of VRAM on these cards

So stay away from the 3080s (especially MSI's Gaming X Trio, that particular card is terrible and nobody should buy it) and wait for the ti and/or Super versions, people

Care to elaborate? This card got reviewed very well.
 
Oct 27, 2017
1,430
This is helping my resolve to hold off on buying a PC until RDNA2 cards are announced. Hopefully it's fixable in drivers but given where they want game development to go in terms of IO it's definitely a bit concerning to see recommendations to under clock.
 

Aeferis

Member
Oct 25, 2017
1,626
Italy
I've spent all day trying to understand why games like Death Stranding are underperforming like crazy to me compared to reviews no matter the settings and with no bottleneck in the system. Got a FE.
 

Pargon

Member
Oct 27, 2017
12,021
As soon as the clock speeds reach a certain level, the game crashes to desktop.
Sounds like a typical unstable overclock to me.
Third-party cards are notorious for this. They try to outdo each other by a few MHz, since many buyers only look for the highest boost clock, and end up with a card which is unstable - if not all of the time, at least in certain games.

People would often blame the game for those stability issues though.
The Evil Within has a reputation for being prone to crashing, while it's almost entirely due to bad overclocks.
After reducing the clockspeed by 50–100 MHz on both a GTX 960 and GTX 970, the game never crashed for me a single time. Before then it was crashing within 15 minutes of loading the game.

A similar thing happened with my GTX 570 in Deus Ex: Human Revolution when tesselation was enabled. In that case, increasing the voltage fixed it.
Since it was one of the early games which used that feature, it would have been easy to blame the game - it was the only one crashing after all.
But it happened because the GPU had shipped with an overclock on the edge of instability that was fine in most games, until tesselation was being used.
DXHR does still have a reputation for crashing though - even when it's not always the game's fault.

And since these GPUs are power-hungry, with a lot of people using power supplies that are just barely meeting spec or technically under it, I wouldn't be surprised if that's a contributing factor as well.

I've spent all day trying to understand why games like Death Stranding are underperforming like crazy to me compared to reviews no matter the settings and with no bottleneck in the system. Got a FE.
How are you determining that there is "no bottleneck" in your system?
You need to be monitoring GPU usage. If it's spending most of its time below ~95% you're bottlenecked somewhere—likely by the CPU.
 

NySpeed

Member
Sep 13, 2020
103
I did happen to install the 1200 dollar zotac from Amazon, and Control did crash to the desktop twice. My 2080 never did that. Anecdotal of course, but I did find it strange. Once one of the fans started clicking on the zotac, I put my 2080 back in and returned it.
 

WetWaffle

Member
Oct 25, 2017
4,604
Not like I had the money to buy one early anyway but will probably delay the purchase by 3-6 months
 

Admiral Woofington

The Fallen
Oct 25, 2017
14,892
Looks like waiting till late 2020/early 2021 was the right call even if I potentially won't be playing cyberpunk right at launch.
 

Mukrab

Member
Apr 19, 2020
7,512
Happens to me in two out of the 4 games im cirrently playing. One is unplayable because it crashes too often, the other one isnt thaaaat bad but still 'playable' i guess
 

Pargon

Member
Oct 27, 2017
12,021
It's spending most of its time below 20-25% lol while my CPU goes up to 50% usage. It's a 3700x so I doubt it can be that much of a bottleneck.
Looking at average CPU usage cannot tell you if it's causing a bottleneck, unless all cores are at 100%.
If it's at anything less than 100% it's indeterminable. You might have a single core hitting 100%, but it's still possible to have a CPU bottleneck without that.

This is why you look at the GPU usage.
If the GPU usage is at less than 100% (though I'd say ~95% is fine) then you are definitely being bottlenecked somewhere.
Note that V-Sync or a frame rate limiter doing its job is also "bottlenecking" the GPU - since that is its purpose. You have to disable those.

Being at only 20-25% seems like there is something very wrong.
Try enabling Debug Mode in the NVIDIA Control Panel (via the help menu) and restarting your PC.
 

Altair

Member
Jan 11, 2018
7,901
Didn't Nvidia have a problem similar to this with the 2000 series with the 2080ti just dying on people when it initially released? Hopefully they can get this reigned in quickly.
 

tokkun

Member
Oct 27, 2017
5,408
I wonder if it has to do with memory temps/stability? There were reports that nVidia originally wanted to go with even faster chips, but then they had to downclock them... that, and I know at least MSI has had a problem with properly cooling memory components in the past.
EDIT: looking at what cards this has been reported on, these two and the FE using the in box power adaptor, it does seem to be a power issue since it's only affecting the 2 power connector card. At least that's my theory.

I won't say it's impossible, but typically a temperature or power issue would result in a full system crash / kernel panic, rather than just an application crashing to desktop. Memory issues also tend to exhibit artifacting.

Application crashes are something you see with unstable overclocks on the core frequency, in my experience. Of course it could be a driver problem too.
 

NovumVeritas

Member
Oct 26, 2017
9,138
Berlin
Didn't Nvidia have a problem similar to this with the 2000 series with the 2080ti just dying on people when it initially released? Hopefully they can get this reigned in quickly.
I've read the same yeah. As an earlier adopter, especially in hardware, I never go the early adopter way, because there can still be issues with software or hardware in general. One of the reasons I'm also waiting.