RTX 3080 Crash

[DarkvaleN]

I also got the same Problem Here with a Ventus 3x Msi. Tested with Hitman 2, Gears 5, Control, AC Odyssey. Even benchmarks freeze after a few seconds.

Curiously 3dMark Timespy or Unigine banchmarks complete without issue, but usually clocks when in those tests hover around 1850-1900.

When the card gets to 1950-2000 MHz Core clock (basically anything that doesn't get  the card to 100% usage), it crashes almost instantly.

 

My suspicion is either driver or maybe the new GPU boost 5.0 feature being too aggressive.

 

I managed to get stable benchmarks at -50/-40 mhz at core clocks.

 

Will try studio driver to see if there's a difference.

 

[Manju1]

1 hour ago, DarkvaleN said:

I also got the same Problem Here with a Ventus 3x Msi. Tested with Hitman 2, Gears 5, Control, AC Odyssey. Even benchmarks freeze after a few seconds.

Curiously 3dMark Timespy or Unigine banchmarks complete without issue, but usually clocks when in those tests hover around 1850-1900.

When the card gets to 1950-2000 MHz Core clock (basically anything that doesn't get  the card to 100% usage), it crashes almost instantly.

 

My suspicion is either driver or maybe the new GPU boost 5.0 feature being too aggressive.

 

I managed to get stable benchmarks at -50/-40 mhz at core clocks.

 

Will try studio driver to see if there's a difference.

 

As usual the question which power supply But i am getting hope that it´s simply a driver issue.

[DarkvaleN]

6 minutes ago, Manju1 said:

As usual the question which power supply But i am getting hope that it´s simply a driver issue.

My power supply is a Corsair TX750M.

 

Nvidia Studio Driver didn't help with the issue in Gears 5 or AC Odyssey. Both crash still at 1440p max settings.

Hitman 2 and Control have stopped crashing after the driver change, which gives me also some hope that is a driver related issue.

 

Will test a bit more, but without a different driver, it's complicated to know if the issue is 100% driver related or not.

[Manju1]

This is what someone mentioned on reddit. I am using one cable with two 8-pins. But i think its bullshit because i found nothing confirming this theorie in the internet (Searched for 8 pin two cables vs one):

 

"As other mentionned, you need to use 2 seperate 8 pins cable (not the 2 connector from the same cable) in order to properly power the GPU. Your 2070 super might not have pulled enough power for you to notice any issues with 1 single cable (I believe it's rated at 215W). But those new cards (3000 series) are power hungry. Usually, 1 x 8 pins can provide around 150W and the PCIe slot provides around 75 - 80W of power. Since the RTX 3080 has been tested to pull up to around 350W under load, you definitely need both 8 pins and a good Power supply to avoid any issues. Your 2070 super was probably OK with 1x8pins + PCIe since it wouldve been pulling 200ish watts in most use case."

 

 

[Deli]

10 minutes ago, Manju1 said:

This is what someone mentioned on reddit. I am using one cable with two 8-pins. But i think its bullshit because i found nothing confirming this theorie in the internet (Searched for 8 pin two cables vs one):

 

"As other mentionned, you need to use 2 seperate 8 pins cable (not the 2 connector from the same cable) in order to properly power the GPU. Your 2070 super might not have pulled enough power for you to notice any issues with 1 single cable (I believe it's rated at 215W). But those new cards (3000 series) are power hungry. Usually, 1 x 8 pins can provide around 150W and the PCIe slot provides around 75 - 80W of power. Since the RTX 3080 has been tested to pull up to around 350W under load, you definitely need both 8 pins and a good Power supply to avoid any issues. Your 2070 super was probably OK with 1x8pins + PCIe since it wouldve been pulling 200ish watts in most use case."

 

Although one cable can carry the current without melting. However the high current that will lead to significant voltage drop, enough to affect stability. jonnyGURU from Corsair has mentioned it in other thread somewhere.

[DarkvaleN]

1 hour ago, Manju1 said:

This is what someone mentioned on reddit. I am using one cable with two 8-pins. But i think its bullshit because i found nothing confirming this theorie in the internet (Searched for 8 pin two cables vs one):

 

"As other mentionned, you need to use 2 seperate 8 pins cable (not the 2 connector from the same cable) in order to properly power the GPU. Your 2070 super might not have pulled enough power for you to notice any issues with 1 single cable (I believe it's rated at 215W). But those new cards (3000 series) are power hungry. Usually, 1 x 8 pins can provide around 150W and the PCIe slot provides around 75 - 80W of power. Since the RTX 3080 has been tested to pull up to around 350W under load, you definitely need both 8 pins and a good Power supply to avoid any issues. Your 2070 super was probably OK with 1x8pins + PCIe since it wouldve been pulling 200ish watts in most use case."

 

 

I was using the 2 8-pin from one cable and I switched to 2 cables, one 8-pin each with no difference.

I had the same thought , but I doubt it is a power supply issue.

 

Thanks for sharing anyway. At the moment I am able to use a -30 Core clock with no crashes.  

[Manju1]

1 minute ago, DarkvaleN said:

I was using the 2 8-pin from one cable and I switched to 2 cables, one 8-pin each with no difference.

I had the same thought , but I doubt it is a power supply issue.

 

Thanks for sharing anyway. At the moment I am able to use a -30 Core clock with no crashes.  

Okay already thought about testing it with two Power supplies (1 for the graphics card one for the Rest of the system) just to verify the solution. But now i will just wait for the next drivers ...

[deadpoor]

2 hours ago, Manju1 said:

This is what someone mentioned on reddit. I am using one cable with two 8-pins. But i think its bullshit because i found nothing confirming this theorie in the internet (Searched for 8 pin two cables vs one):

 

"As other mentionned, you need to use 2 seperate 8 pins cable (not the 2 connector from the same cable) in order to properly power the GPU. Your 2070 super might not have pulled enough power for you to notice any issues with 1 single cable (I believe it's rated at 215W). But those new cards (3000 series) are power hungry. Usually, 1 x 8 pins can provide around 150W and the PCIe slot provides around 75 - 80W of power. Since the RTX 3080 has been tested to pull up to around 350W under load, you definitely need both 8 pins and a good Power supply to avoid any issues. Your 2070 super was probably OK with 1x8pins + PCIe since it wouldve been pulling 200ish watts in most use case."

 

 

This is a relative new problem, I found some info and half is chinese.

A 2018 discuss and jayz 2cents 

MSI engineer Toppc Lin talk about temp different with different number of connector.  

 

 

Seasonic installation guide. https://knowledge.seasonic.com/article/8-installation-remark-for-high-power-consumption-graphics-cards

 

 

A guide of connector on Chinese PSU review site FCPOWERUP. https://www.fcpowerup.com/cpu-8pin-4pin-connector/

[Manju1]

Mhm strange thing still is that downclocking by -50 resolves the issue. I am definitely not an expert and it still might be the power supply, but -50 MHz downclock doesent seem so much less that the power supply is suddenly enough for the card.

[YandereX]

 

7 minutes ago, Manju1 said:

Mhm strange thing still is that downclocking by -50 resolves the issue. I am definitely not an expert and it still might be the power supply, but -50 MHz downclock doesent seem so much less that the power supply is suddenly enough for the card.

I agree with you, -50 from my pov is too minimal of a number to be a power supply issue. I imagine if the card was really hitting a power ceiling there'd be more then just the game crashing. Such as a Windows BSOD or a Driver Crash.

In my case of using a EVGA SuperNOVA 750w E+ The point in which my card is stable is 2025, From 2040 which seems to be the deadzone. -15 shouldn't quite be the point were the card turns over. Though at the same time the Studio driver seems to be a lot more easier to run then the game ready driver, Being that the studio driver runs my card at 1985-2025 while the GRD runs it at 2025-2040 with some instances of going to 2055
 

Card: MSI Gaming X Trio RTX 3080

[zarthere]

My RTX 3080 GAMING X TRIO has this issue too. When I play demanding games (RDR2, FC5, MW and some other games) the game freezes and then crashes me to my desktop. This usually happens within 20 minutes. During gaming I sometimes get a mini one second freeze as well (in demanding games).

 

When I look at my windows event log I see that every time I get the crash, a log file is made, showing the error: ''Display driver nvlddmkm stopped responding and has successfully recovered.''

 

When the crash happens, sometimes I get an error pop up. In RDR2 I always get a ''ERR_GFX_STATE '' error. And in Modern warfare I get the ''Dev error 6068, directX encountered an unrecoverable error''. In the other games my screen just freezes as well and then crashes to my desktop without a specific error.

 

So far, non-demanding games seem to work. I tried WoW and did not experience any issues, but the game is not demanding and puts my GPU only on 40% usage. I also tried all 3D mark benchmarks and these run fine, but I noticed the clock speed during those benchmarks is lower than during actual gaming sessions. 

 

I checked my temps and they are all fine (GPU gets to about 70Celcius) When I underclock the clockspeed of my GPU by 80Mhz, It seems like I don't get the crashes anymore, but sometimes I do still get 1 second freezes. The factory boost speed reaches up to 2040Mhz but apparently this is unstable as downclocking it by 80Mhz gets rid of the crashes. However during games I still sometimes get one second mini freezes (stuttering) 
 

I wonder if we got early adapter bad batch cards or if this is a driver issue..

[YandereX]

On 9/22/2020 at 9:55 PM, zarthere said:

I wonder if we got early adapter bad batch cards or if this is a driver issue..

Honestly at this point, I keep telling myself this is a driver issue as this isn't the first time i've had this issue with a graphic card. This is how it was when i purchased a 1070 back when it was released and eventually with further driver updates it got better with time. I'm vouching on this is probably the same situation here. Especially with the way so many brands are having the same issue. Like it'd be fair if it was like "Well MSI Cards are only having this issue" but to see Gigabyte,Asus,Msi, Zotac cards have the same exact issue has me feeling suspicious that either A, The PCB that AIBs received were half scuffed and half good, Or B, The Driver itself is the issue and from what i've seen this driver [456.38] has been a nightmare for many who don't even have the 30 series cards.At the very least the Studio driver in most games excluding, RDR2 as i get less frames on that works fine and i haven't had an issue with games like Rainbow 6 Siege, Hopefully it settles itself soon though 

 

Edit: Red Dead is running fine and the frames are good too, Will update back if it crashes on the Studio Driver

Edit 2: Red Dead worked for hours on end unlike the GRD but ultimately still crashed

[DarkvaleN]

9 hours ago, YandereX said:

 

I agree with you, -50 from my pov is too minimal of a number to be a power supply issue. I imagine if the card was really hitting a power ceiling there'd be more then just the game crashing. Such as a Windows BSOD or a Driver Crash.

In my case of using a EVGA SuperNOVA 750w E+ The point in which my card is stable is 2025, From 2040 which seems to be the deadzone. -15 shouldn't quite be the point were the card turns over. Though at the same time the Studio driver seems to be a lot more easier to run then the game ready driver, Being that the studio driver runs my card at 1985-2025 while the GRD runs it at 2025-2040 with some instances of going to 2055
 

Card: MSI Gaming X Trio RTX 3080

 

From my tests, the card it's still drawing the same amount of power, around 320W at 3dMark, so it cannot be power draw.

A simple test is to limit power to 90 or 80% in MSI afterburner, but leave the clocks untouched. You will see the power draw go down to around 280W, but it will still crash.

 

A simple change from GRD drivers to NSD drivers made crashes in some games disappear, without doing anything else, so it seems reasonable to think it affects somehow.

 

 

[Alex.J]

Yeah I'm getting the exact same problem as a lot of people on this thread with my MSI 3080 ventus 3x OC where it crashed out of games back to the desktop. From what I can see it's when the gpu hits around the 2000mhz+ mark it crashes back to the desktop. The only fix I found is to downclock too by 50mhz or so. Hoping this is a fixable driver issue. This happens the fastest on bfv as soon as it gets to the menu it's crashes 

 

SPECS

PSU:Corsair TX750m 

MOBO: MSI pc pro

GPU: MSI 3080 ventus 3x OC

CPU i7 8700k delid 5ghz 1.35v

Ram: Corsair 16gb 2400mhz 

[Manju1]

Feels a bit strange that nvidia is deleting the posts on the offical reddit 

 

 

[Danoniero]

A lot of people here and there reporting that plugging 2 separate 6+2/8pin connectors instead of daisy chaining solved the issue.

[Apington]

Just replaced my Corsair AX850 Titanium (bought for this new lovely gpu - MSI Ventus 3X 10G OC) with a beQuiet Straight Power 11 1200W Platinum.

 

Positive:

- far less coil whine

- 3D Mark got me some points (12 928) ( https://www.3dmark.com/3dm/50743594?

- 3D Mark stress test was fine (98.6%) https://www.3dmark.com/3dm/50746602?
- temps are good (71 °C)

 

Sadly the negative points are still the same... almost any game I try to launch dies (CTD) within the first 1-2min at it's best (e.g. Borderlands3, Warzone crash right after loading screen).

As stated before I can only hope that a new driver can fix this. The benchmarks just seem too good for a defect chip - at least I guess so.

 

Slowly you can find more and more voices / threads about 3080s not running any games - https://wccftech.com/nvidia-geforce-rtx-3080-users-report-crashes-black-screens-during-gaming/ (the last summary I found)
 

If anybody has suggestions or questions feel free to ask, I will provide whatever info I can.

PS: Without any fixes for the next week or so I will have to return it to my reseller. Not paying 700€+ for non functioning hardware.

[Mark Kaine]

that sounds like a possible cable issue. it would definitely be better to use 2 separate cables instead of just one with the tacked on second connector... that seemed to be a problem from the very beginning, Nvidia should probably have some instructions included to use 2 separate pcie cables (if that really is the issue)

[zarthere]

53 minutes ago, Danoniero said:

A lot of people here and there reporting that plugging 2 separate 6+2/8pin connectors instead of daisy chaining solved the issue.

Most people I talked to already have separate cables (including me) but this does not solve the problem. I did a lot of testing and I am pretty confident that in most of the cases, the PSU is not the issue. 

[zarthere]

31 minutes ago, Apington said:

Just replaced my Corsair AX850 Titanium (bought for this new lovely gpu - MSI Ventus 3X 10G OC) with a beQuiet Straight Power 11 1200W Platinum.

 

Positive:

- far less coil whine

- 3D Mark got me some points (12 928) ( https://www.3dmark.com/3dm/50743594?

- 3D Mark stress test was fine (98.6%) https://www.3dmark.com/3dm/50746602?
- temps are good (71 °C)

 

Sadly the negative points are still the same... almost any game I try to launch dies (CTD) within the first 1-2min at it's best (e.g. Borderlands3, Warzone crash right after loading screen).

As stated before I can only hope that a new driver can fix this. The benchmarks just seem too good for a defect chip - at least I guess so.

 

Slowly you can find more and more voices / threads about 3080s not running any games - https://wccftech.com/nvidia-geforce-rtx-3080-users-report-crashes-black-screens-during-gaming/ (the last summary I found)
 

If anybody has suggestions or questions feel free to ask, I will provide whatever info I can.

PS: Without any fixes for the next week or so I will have to return it to my reseller. Not paying 700€+ for non functioning hardware.

May I ask which 3080 you’ve got and how bad the coilwhine was and how big of a difference the new PSU makes for the coilwhine? My X TRIO has pretty noticable coilwhine and im thinking about replacing my 1 year old 850W platinum corsair PSU with a 1000W one if it helps against coilwhine. 
 

Also, I’m gonna wait 1 week for a fix or statement as well. If it hasn’t been resolved by then I am also returning my card. Its an odd situation we’re in really. 

[Apington]

22 minutes ago, zarthere said:

May I ask which 3080 you’ve got and how bad the coilwhine was and how big of a difference the new PSU makes for the coilwhine? My X TRIO has pretty noticable coilwhine and im thinking about replacing my 1 year old 850W platinum corsair PSU with a 1000W one if it helps against coilwhine. 
 

Also, I’m gonna wait 1 week for a fix or statement as well. If it hasn’t been resolved by then I am also returning my card. Its an odd situation we’re in really. 

Got the MSI GeForce RTX 3080 Ventus 3X 10G OC.

With the Corsair AX 850 the coil whine was pretty disturbing. Could realy hear it ramp up, even with case fully closed. Now I maybe heard some slight, but with open case and a silent room.

Still it mostly occures when the card ramps up.

[zarthere]

20 minutes ago, Apington said:

Got the MSI GeForce RTX 3080 Ventus 3X 10G OC.

With the Corsair AX 850 the coil whine was pretty disturbing. Could realy hear it ramp up, even with case fully closed. Now I maybe heard some slight, but with open case and a silent room.

Still it mostly occures when the card ramps up.

Oke thanks man, so sounds like a decent improvement in noise. I’ll think about getting a higher wattage PSU. 
 

but first,  hopefully the issues with the crashes get fixed with a new driver. Im at least expecting a statement from Nvidia and/or the AIB partners about this matter. As its driving me nuts to not know whats going on with my GPU. 

[Elcid_888]

Hey guys, there are several threads and many, many people on Reddit complaining of the same issue. This is widespread. Check out my link.

 

Nvidia is not addressing the issue either. Ridiculous. 

[Quickstrike]

On 9/18/2020 at 7:40 AM, sindain said:

HI,Please Help

I just bought a new powersupply and a new rtx 3080 . Whenever I play games it keep crashing ever 3 minutes or soo
I have no idea whats the issue. I have reinstall the nvidia driver and it still keep crashing.

I also found out that my power supply have some sort of whine sound. 

but I just bought it new for less than 3 days

Please help

this is my spec
 

• Windows 10 64 bit
• ROG Z370-H
• Intel i7 8700k (Do not over clock)
• 2x8GB Corsair Vengeance RAM 
• RTX trinity 3080 Zotac 
• WD blue ssd 1tb 
• Corsai CX750M

Thank you

Are you using a PCI-E riser cable?

 

[boasty]

It's not the memory modules getting too hot under boost clock?  Card temperature might be showing ok but perhaps the memory is getting too hot. There have been tests done with the ddr6+ memory on the 3080 and some of the chips were getting over 100 degrees. Might be enough to create a memory error and crash to desktop. The article I read about the issue:-

https://www.igorslab.de/en/gddr6x-am-limit-ueber-100-grad-bei-der-geforce-rtx-3080-fe-im-chip-gemessen-2/

 

Hopefully it's not that and just a driver issue but I think it's a useful article and worth knowing about..