Diablo 4 Bricked my GIGABYTE 3080 TI

Technical Support
142

For me I played the last beta using an ASUS ROG STRIX G513QY laptop. I have upgraded the RAM from 16Gb to 32GB, and added an extra PCIe3.0 SSD which is the one the game is installed on. The CPU is the Ryzen 9 5900HX (Factory OC) and the GPU is the Radeon 6800M 12GB.

I was running December 2022 drivers (because I had not used my laptop in a while). After about 3 hours of playing with all detail set to 1440p Ultra (auto selected), I got a black screen, but the laptop’s thermal system did not ramp up. I was a bit concerned and started trying the usual things of Alt Tab, ctrl alt del, alt f4, etc. Nothing worked, so I was about to reach for the laptop’s power button as it was to the side and I was playing on a 32" 1440p 144Hz monitor with normal keyboard and mouse.

Right as I started to reach for the power button, the screen game back (it was off long enough for the monitor to enter sleep mode, at least 10-15 seconds, which is longer than a driver reset), there was a Diablo error message, the game was still open but I was disconnected from my session and an error which simply stated that something went wrong. I closed the game, and then noticed I was on December 2022 drivers and figured maybe I should update those. I also looked on the ASUS website and saw a VBIOS update which I had not flashed yet.

After updating the VBIOS and the drivers to the latest, I played the D4 beta until it closed down for hours and never suffered any other issues. The most I had was some basic stuttering which I attribute to worse than normal latency as I am in Australia and I assume the beta servers were in the USA giving me automatic 150-200ms ping. And the textures may not be fully optimised yet as the game is in beta.

So for me, that was a completely different hardware loadout, and similar issue that did not result in anything being bricked.

I think the people in this thread who are blaming the game only, are incorrect. The game itself cant possibly kill a GPU hardware. But the game can push a GPU hard, if the GPU is a good brand with super high quality construction and the user’s other system hardware is equally high quality, there shouldnt be an issue.

I personally run all my electronics from a Double conversion UPS (Powershield Centurion RT 2000va). It provides clean AC voltage and frequncy to my electronics even when sometimes on hot days in our area the input voltage registers at 224v @ 43Hz (normal is 240v @ 50hz), the UPS always counters it and provides the system clean stable 240v @ 50hz.

I recently had an ASUS THor 1200W Platinum PSU fail its 3.3v rail, and that happened suddenly, and the system only ever ran from the UPS, so it doesnt matter if you buy the $1000 PSU or the $200 PSU, they can fail and can damage hardware. Just like a game can send a set of instructions to a GPU that creates the perfect scenario where the card asks momentarily for astromonical amounts of power and either trips some form of OCP or damages the card.

Is that the game’s fault? No, its the hardware vendors fault for not building the card good enough to sustain the transient spikes it can generate, or not limiting the card as so it cannot generate spikes that will kill the card’s power delivery system.

Vendors may not want to limit the cards performance to stop it from creating spikes in energy requirement that can kill sub-par power delivery systems. Why? Because to do so would make thier card noticeable worse performing than the competition, so they choose to not limit the card and hope that no cards are put in a situation where they cant cope and fail. That is not the fault of a software developer, that is poor form of the hardware vendor.

1 Like
143

Gigabyte are a terrible company. I recently built a new PC and avoided them like the plague.

3 Likes
144

The enthusiast world is getting harder to live in; I had an issue with Asus where they didn’t bother to pick a good drive for a line of their laptops then lost a class action case due to this, so they’re trash too. Gigabyte’s out, EVGA is no longer making GPUS, now what.

145

Hello everyone,

The Diablo 4 team has been investigating the reports of GPU issues mentioned in this thread and elsewhere online. The team is working closely with Nvidia to identify affected hardware configurations and gather as many data points as possible to assist in the investigation.

If your card is still not working, try these troubleshooting steps:

  • Conduct a complete power cycle (AC OFF->AC ON) and see if GPU can be detected under Windows.
  • If GPU can be found under Windows, do a clean driver install.

If that doesn’t work, please reply and include the following information or logs:

  1. GPU card vendor and model number
  2. Display resolution when running Diablo IV
  3. Display mode (Windowed/Windowed Fullscreen/Fullscreen)
  4. Max Foreground FPS, Max Background FPS
  5. Diablo IV graphic quality settings, preset with Low/Medium/High. Or more detail with custom settings.
22 Likes
146

Luckily, my card was not bricked but I was still getting a crash to desktop everytime I went into any dungeon. The only error I received was it saying “The game has encountered and unexpected error.” Then gave me a random string of numbers and letters that were different everytime the game crashed.

Info on PC:
1 - Asus ROG Strix 3080ti
2 - 1440p (I also tried 1080p but the issue still persisted.)
3 - Windowed Fullscreen (I tried all display mode options but the issue still persisted.)
4 - Max foreground FPS was 165 and background was 60 (I changed this to different settings as well - 144, 120, and 60 FPS. Issue still persisted.)
5 - High settings ( I chaned my settings trying everthing from High to Low and the issue still persisted.)

I can’t remember any specific crashes being in the overworld. The majority (if not all) of my crashes to desktop happened in dungeons. Some were story dungeons, some were random ones on the map.

Would love to see a fix for these issues.

147

My card didnt die, but I had a black screen + full power fans, so I’ll post the details. It happened after long afk(like 40 minutes) - can’t remember exactly if it was a random dungeon or first big town kyovashad

  1. NVIDIA GeForce RTX 3080 Ti by Gigabyte (didnt overclock it) 531.26 driver
  2. 1440p
  3. Windowed Fullscreen.
  4. Max foreground - 143, max background - 143. Was also locked at 143 via Rivatuner
  5. High settings, I changed some of them to medium after the black screen.

Also I noticed some weird behaviour with alt-tab - sometimes game get frozen after alt-tab until I alt-tab one more time.

148

No crashes on my side. I’m not actually using a 3080ti; I have a RTX 2060 but I still want to provide some feedback regarding performance because it’s tanking on the same portions that 3080ti had crashes.

I experienced significant performance issues during most cutscenes, particularly at the chapel cutscene and both cutscenes with Lorath, which caused temperatures going from 60C to >85C and fans spinning at a faster speed; though no max speed. I also had a lot of stutter and FPS drops in certain portions of the map.

  1. GPU: NVIDIA GeForce RTX 2060
  2. Resolution: 1080x1920
  3. Display Mode: Windowed Fullscreen
  4. Max Foreground FPS: 120 (tried at 60 too but didn’t seem to have any significant changes), Max Background FPS: 8
  5. Graphic Settings: I tried various combinations. The one that appeared to be most unstable was a Custom Medium (tweaked a few things and disabled heavy particle effects, but kept shadows on ‘High’ for more immersion).
1 Like
149

Did not brick but would error out randomly to Fenrir. In the three days of beta I was still stuck in prologue with the reconnects.

  1. 3060 Ti Gigabyte
  2. 1080p
  3. All three display modes attempted
  4. Max foreground was the only thing I didn’t lower - 150. Ran out of beta to test, as Furmark has the card as it sits in system over 170@1080p.
  5. All graphic settings from high to low with custom removing all “… This will affect performance…” removed
150

My card (Gigabyte 3080ti Vision OC 12GB) did not brick, luckily, but it did stop sending a signal to the monitor (was using default graphics settings, including Fullscreen). Prior to that the fan usage was fine, immediately when the signal was lost fan usage was at 100%.

This did not occur during a cutscene, but rather a specific spot in the open world near Father’s Cross. The first time there was a mob of fallen with at least 1 “glimmering” fallen. The second time I could hear fallen, but could not see them. In both instances the sun’s reflection off the snow seemed incredibly bright.

This was the only place where I encountered an issue, prior to the second issue I had loaded into the map closer to the Kyovashad gate and was able to progress in the city and points North; it was only when I returned to trek through the southwest gate of Kyovashad again to get the waypoint in Nevesk that I encountered it. Both events occurred after roughly 1.5 hrs. After the second time, I stopped playing not wishing to tempt fate.

In both instances within seconds I held in the power button of the system until it turned off and then turned off the PSU.

151
  1. GPU card vendor and model number: Gigabyte 3080 Ti Vision OC ‎| GV-N308TVISION OC-12GD
  2. Display resolution when running Diablo IV: 1440p
  3. Display mode (Windowed/Windowed Fullscreen/Fullscreen): Windowed Fullscreen
  4. Max Foreground FPS, Max Background FPS: 150, 30 (Default settings)
  5. Diablo IV graphic quality settings, preset with Low/Medium/High. Or more detail with custom settings.: Just used “High/Ultra” settings. Nothing special.

My card was 100% bricked and I just sent it back for RMA today to Gigabyte. Who knows how that RMA process is going to go…all I’ve heard is bad things.

152

what resolution, what fps?

153

In this entire thread of 150+ comments, only 2 or 3 people claimed to have a bricked GPU (with nothing confirmed, just a claim, no proof) and this thread has been shared on many major sites including reddit, Diablo community sites, etc.

Do you have a source for that?

154

my gpu bricked too :frowning:

1 Like
155

It is 100% your gpus, and poor production of them.

Gigabyte and some Asus. Mainly 3080 TIs.

The onesy twosey other random cards, arent a trend, its a coincidence, and their cards were faulty.

Heres a video for you. Most saying “its the game duh” wont understand it, but try.

https://www.youtube.com/watch?v=or7njUlYTLI

For those too dense to understand all of it, skip to and watch 10:30 to 10:40.

Its the manufacturing of the card, and issues there. The game simply exposes those issues. Down the road, another game would do the same thing.

Contact the manufacturer, (good luck right now lol)

A game cannot and will not brick a card thats built solid. Think about it for two seconds. If it was the game, why would it only be bricking select cards, and not all of them, ESPECIALLY the lower end cards?

Exactly.

1 Like
156

mine “recovered” every time
had total 15-20 crashes during weekend
HW: HX1000 PSU + GV-N308TGAMING OC-12GD
FENRIS LOG showed: DXGI_ERROR_DEVICE_RESET or DXGI_ERROR_DEVICE_REMOVED

latest drivers, updates [system - w11, mb or gpu bios] and stock gpu settings

the issue was: black screen or some artifact + 100% gpu fans - had to restart the pc

happened on both max and mid texture settings [more often on max]
max texture quality also caused very heavy stuttering of the game

full cfg if that helps:

cpu: Intel Core i7 11700KF 8x3.6 GHz BOX (LGA1200-G11, 16MB, 125W)
gpu: GIGABYTE GeForce RTX3080TI Gaming OC, 12GB, DDR6X, PCI-E, 3xDP, 2xHDMI
mb: Gigabyte Z590 Gaming X, Intel Z590, DP, DDR4, SATA3, 3xM2, RAID, ATX, LGA1200-G11
ram: Ballistix MAX RGB, DDR4, 32 GB, 4400MHz, CL19 (BLM2K16G44C19U4BL)
ps: ATX 1000W Corsair HX1000 (135mm, Platinum)

one more thing noticed in event logger:

Application Popup 56
ACPI 2

157

So there it’s fault on both sides - blizzard and gigabyte, no other game pushed my 3080ti into black screen(luckily survived) in almost 2 years of intensive gaming. Yeapp. you could say that those gpu were bad, but still other games were unable to push them at such power level to brick it. Anyways, beta was patched, nividia released new drivers, I’m 90% sure that issue is already solved, but gonna wait for a day for further reports.

P.S.
I guess some damage control from blizzard is going on here, not to hurt the sales, huh.

1 Like
158

Its poor voltage regulation iirc. A good card will protect itself. Any good hardware will protect itself. If you have a car, and its brakes fail due to being built poorly, are you going to blame the road its traveling on? No. You aren’t. You’re going to take it right back to the manufacurer and tell them to fix their crap product.

2 Likes
159

Nvidea just relesed new deiver as of march 23, 2023. Update everyone.

2 Likes
160

If you were able to drive that care for years before, and suddenly you get into a road that’s built of broken bricks - I’ll blame the person who built that road.

161

But that road, if it was the issue, would be destroying all cars, not just 1 or 2 select models… proving those models are poorly built, and are the problem.