EVGA RTX 3090 FTW3 Ultra Troubleshooting Help

So I turned off curve optimizer but did not turn off PBO.

Tried 2 games. Dirt 5 crashed with this error.
Faulting application name: game_release.exe, version: 0.0.0.0, time stamp: 0x6029c8b4
Faulting module name: VCRUNTIME140.dll, version: 14.28.29231.0, time stamp: 0x5f4c66c9
Exception code: 0xc0000005
Fault offset: 0x00000000000012de
Faulting process id: 0x1494
Faulting application start time: 0x01d70d16bd7f73d4
Faulting application path: C:\Program Files\WindowsApps\CodemastersSoftwareCompan.DiRT5_1.2579.81.0_x64__4cfye3zbe1gaw\game_release.exe
Faulting module path: C:\Program Files\WindowsApps\microsoft.vclibs.140.00.uwpdesktop_14.0.29231.0_x64__8wekyb3d8bbwe\VCRUNTIME140.dll
Report Id: 61fc343b-6253-452b-a606-f9d5c81bd250
Faulting package full name: CodemastersSoftwareCompan.DiRT5_1.2579.81.0_x64__4cfye3zbe1gaw
Faulting package-relative application ID: game.release

And then I tried Immortals Fenyx Rising and it also crashed.

Next up is to turn down ram overclocks and see what's up. This is extremely annoying. I had no such issues without the 3090. :hmm:
 
Are your memory sticks getting heat dumped on them by the card? Some memory types are temperature sensitive. I'm always running high memory overclocks so I have to not only use direct cooling on the sticks, but also redirect the GPU heat away from them. Anyway, more GPU power = more heat = memory issues?

If you have a box fan or a table fan you could try placing it so it blows on your memory and see if that helps. Also you could up the memory/SOC/VDDG voltages a notch (assuming you're not running them high already).
 
So I backed down the overclock on CPU and was very happy because everything was running fine. I also set memory back to default XMP profile. The only thing overclocked is FCLK at 1867 and DDR4 at 3733. I am going to back that down now as well.

Played Immortals Fenyx Rising for more than 1 hour. Then logged onto the Dirt 5 and everything was fine until got a crash yet again. My theory is that CPU PCIe lanes can't handle the load and I need to back down FCLK to stock otherwise, the card is causing the crash. Something I have no way to prove since all crashes seem to be software related.

Rebooting now again. :nag:
 
Well I reset bios and only set XMP to Profile 1 (DDR4 3600) and everything else on Auto and Dirt 5 still crashed. It seems this is a common issue.

Wtf all games I am playing are crashing and it is a common issue? :nuts: You can search Dirt 5 crashing to desktop and there are tons of users on steam or gamepass having this issue not only on PC but also on PS5 and XSX. :nuts:

Time to try out COD CW and Immortals Fenyx Rising. Immortals I played for quite a lot without crashes. Now my PC is fully bone stock. I am just pissed off.
 
Not really. I got the latest driver. I can maybe go back to the previous one. Drivers are different this time around. That is correct since a new one was released 2 days ago.

What a disaster. I spent so much time trying to replicate issues since they don’t crop up immediately. For example right now my son has been playing CW with DLSS for past 30 minutes and no crash. Wtf is this? I am literally losing half hour to 1 hour before an issue is reproduced. Will try new driver tomorrow morning now.

:runaway:

1 hour 22 minutes of COD BOCW w/ DLSS. No crashes. :nag: :nuts:
 
Last edited:
KAC said:
Not really. I got the latest driver. I can maybe go back to the previous one. Drivers are different this time around. That is correct since a new one was released 2 days ago.

What a disaster. I spent so much time trying to replicate issues since they don’t crop up immediately. For example right now my son has been playing CW with DLSS for past 30 minutes and no crash. Wtf is this? I am literally losing half hour to 1 hour before an issue is reproduced. Will try new driver tomorrow morning now.

:runaway:

1 hour 22 minutes of COD BOCW w/ DLSS. No crashes. :nag: :nuts:
Click to expand...

Well KAC I've got an AMD 3800X and a KFA2 3090 and I've played a lot of Dirt 5 and COD CW and never had a single issue. I previoulsy had an AMD 5700XT and all I did was nuke the AMD drivers with DDU in safe mode and installed the new 3090 and bingo no issues. Haven't had to reinstall windows and all benchmarks I've run with everything stock are where they should be. I'm also running everything on an FSP Hydro 850w PSU.

IIRC when the 3080/3090 were launched there was an issue with some AIB's cheaping out on capicators which was causing the type of issues you are experiencing. EVGA admitted to being one of them. The full story is here.

https://www.techspot.com/news/86900-capacitor-issues-causing-rtx-30803090-crashes.html

Unfortunately whilst you might hope it's a software issue it's most likely a GPU issue and you'll have to bite the bullet and RMA it. Sometimes the answer is staring you right in the face.
 
The caps are a non-issue and EVGA uses 4 POS/2 MLCC. That entire thing was overblown over one dumbass who began the rumor, and then other media outlets ran with it. There are 5 POS cards that are better clockers than the STRIX which has 6 MLCC.
 
Nunz said:
The caps are a non-issue and EVGA uses 4 POS/2 MLCC. That entire thing was overblown over one dumbass who began the rumor, and then other media outlets ran with it. There are 5 POS cards that are better clockers than the STRIX which has 6 MLCC.
Click to expand...

If you google EVGA and 3090 problems there are a lot of reddit and EVGA forum posts about failing cards. All I'm saying is it's more likely to be a defective card than a software problem whether KAC likes it or not. When I had my issues with Vega 64 and my PSU I balmed everything and tried every fix in the book until eventually I had to accept it was a PSU issue.

KAC you could try and undervolt the card and there is a guide on Igor's lab on how to do it. That should at least rule out power or a spiking issue. At least it won't do any harm.

After that good hunting.
 
KAC could just have a bad card too.

IMO, if there are issues with the card after a clean install of windows and on several different games within the first week of use then it's time to RMA it.

Sounds like a bad card to me.
 
The crashes are not related to the GPU. The error he posted earlier looks like something caused by a WHEA or CPU/mem instability.

I'd try rolling back one driver set just to be sure. Besides that, I guess RMA the GPU because that's the only thing that changed since the system was supposedly stable. I don't really know what else you can troubleshoot at that point.

I'm not saying the card can't be bad, but it's certainly not due to the POS/MLCC capacitor crap that the weird dude pushed for hits on his youtube channel. He's the same dummy who claimed you could flash the FTW3 bios to an XC3 for more performance, but made sure the "you must shunt mod the card" part was left out.. in which case, the BIOS doesn't ****ing matter :lol:
 
If he had another video card he could pop it in and see if it continues to toss out errors.

If it does then it's more than likely his setup.

If it doesn't then I'd RMA the card.
 
Today is supposed to be work day but here are my plans to try different things. Computer and card are being run bone stock now so there is no OC causing instability issues. Not in this order of course.

1) roll back a driver
2) try Dirt 5 with 3070
3) remove power extensions (I wanted to do this when the hybrid kit arrived but now I won’t wait)
4) try to install one more graphically intensive game to see if it is a card issue (maybe control) since I never had a single crash in that game like ever

If I do still get crashes then I will have no choice but to try and get an RMA. It will be quite an experience since I am sure EVGA will rape me on shipping costs. Oh well.
 
A good idea would be to get everything as stock as possible to isolate the video card. That would include direct wiring. Basically, you want to isolate only one variable if possible.
 
You must log in or register to reply here.
Back
Top