RTX 3080 Crash

[Apington]

Fyi, just found this on nvidia subreddit:

https://www.igorslab.de/en/what-real-what-can-be-investigative-within-the-crashes-and-instabilities-of-the-force-rtx-3080-andrtx-3090/

[Manju1]

Sounds not good for me O_o My old Graphics card is already gone and if i need to replace the RTX 3080 it i will have no Pc for quite some time xO

(MSI Ventus 3x RTX 3080)

 

Just as an status update: 

I switched from my thermaltake 700w PSU to an Corsair RM850x. You might have guessed it... nothing changed. Still crashing after 5-15 minutes of playing (Without the downclock of 50 MHz).

[MrPurple84]

On 9/18/2020 at 2:02 PM, Mykull said:

Are you plugged into multiple surge protectors, or do you have to many devices plugged into the same surge protector as your computer.

 

try unplugging everything but your computer and monitor/the bare essentials if the above is the case. My computer kept crashing, my 3950x would under clock and stayed at 1 ghz, i had frame drops, and often times I was only getting 20 fps and had weird power supply issues, then I decided to plug less things into my surge protector that was connected to my computer, and well the problem went away. I think it had to do with overall power draw from the outlet. Basically my computer was not getting as much electricity because it was going to my other devices that were plugged in.

 

you can also try a UPS that produces a sine wave or “clean energy” for your computer to see if that makes it run better

This fixed it for me. 

Plugged PSU straight into wall outlet, and that stopped the crashing for me, everything is great now. 

 

Zotac 3080, I9900k,32gb RAM, H150i, Corsair RM750w psu.

 

Firestorm program is shit though.

[NinjaChemist]

On 9/18/2020 at 7:55 AM, YsGrandi said:

hardware unboxed review with ryzen 9 3950x and 3080 pulled 520 watt so even a 650 watt psu is enough

1000watt is just wasted money for single gpu setups

what more important is the PSU quality

he would have been better off with a 650 tier A psu than this low tier 750 watt corsair

 

I don't like to tell people how to spend their money. It's not "wasted" money. There are reasons to buy a beefy PSU. They're most efficient at about half power. I bought a 750W unit a few years ago, using the 'gaming/NAS' video as inspiration. I wanted a low draw because I planned to leave it on all the time. My plex server idea never panned out, but I have enough juice to run my 3080. It all worked out in the end.

[Deli]

1 hour ago, NinjaChemist said:

I don't like to tell people how to spend their money. It's not "wasted" money. There are reasons to buy a beefy PSU. They're most efficient at about half power. I bought a 750W unit a few years ago, using the 'gaming/NAS' video as inspiration. I wanted a low draw because I planned to leave it on all the time. My plex server idea never panned out, but I have enough juice to run my 3080. It all worked out in the end.

But the efficiency curve of a gold rated PSU is 89% at 20% load, 92% at 50% and 90% at full load. Even it isn't right at the peak of 92%. Likely it's still operating at about 90%. The difference of 92% and 90% doesn't seems a big deal. People keep justifying buying big wattage PSU so it will operate at peak efficiency is ........

[zarthere]

3 hours ago, MrPurple84 said:

This fixed it for me. 

Plugged PSU straight into wall outlet, and that stopped the crashing for me, everything is great now. 

 

Zotac 3080, I9900k,32gb RAM, H150i, Corsair RM750w psu.

 

Firestorm program is shit though.

What issues did you have before trying this? Did you crash in games unless you downclocked? 

[Puss]

14 hours ago, zarthere said:

Also, when I measured usage and stability I saw wattage sit stable between 320- 330W and the voltage was switching every few seconds or so between 1V and 1.1V. Is this < 0.1V switch not normal? Or is this actually an issue and not supposed to happen at all, making this 0.1V switches cause crashes?
 

EDIT: Oh I see that you said LOW EFFICIENCY PSU’s can be a fix aa they use different components. Wouldn’t that mean the issue lays within the power stages of the GPU, as the PSU’s are fine? 

You can't see voltage stability without probes on the PSU outputs and the PCIe connector. 0.1V switches should be normal, but self reported values from built in probes generally won't show power quality issues.

 

Yes. These cards sound like they need larger capacity capacitors, but....

 

No. The PSU's aren't fine. In the quest for efficiency, power quality has been ignored, because up until last week, it wasn't important. If harmonics can cause interference (MrPurple84 said removing everything on the same power board solved their issue), this is a power quality issue, with efficient PSU's not being able to supply a clean enough or stable enough voltage.

[Gorou92]

4 hours ago, NinjaChemist said:

I don't like to tell people how to spend their money. It's not "wasted" money. There are reasons to buy a beefy PSU. They're most efficient at about half power. I bought a 750W unit a few years ago, using the 'gaming/NAS' video as inspiration. I wanted a low draw because I planned to leave it on all the time. My plex server idea never panned out, but I have enough juice to run my 3080. It all worked out in the end.

that because you know what you want

but for someone asking here most likely he won't be using that much power and may be wasting it on idle load and be more inefficient

[sindain]

Hi guys 

An update here 

I have fix the issue by downclocking the card by -50

I have found out that is zotac issue because it cheapens out on the component.

I have returned my card since i dont want to have a problem with my card in long run 

 

I have replace the my card with asus rtx 3080 tuf and since then it got no issue whatsoever.

 

 

My suggestion go return your card since is the hardware issue

[zarthere]

5 minutes ago, sindain said:

Hi guys 

An update here 

I have fix the issue by downclocking the card by -50

I have found out that is zotac issue because it cheapens out on the component.

I have returned my card since i dont want to have a problem with my card in long run 

 

I have replace the my card with asus rtx 3080 tuf and since then it got no issue whatsoever.

 

 

My suggestion go return your card since is the hardware issue

Thanks for sharing, yep indeed seems like its now pretty much confirmed to be a hardware issue. However its not just zotac, almost all AIB’s have made the same mistake here. Have fun with your TUF, seems to be a great card! 

[MrPurple84]

I also had issues using the adapter cables, could not find the card, just the lights would power on. Used a 2pin to go with the 6 and 8 pin connector i already had, and it worked fine.

 

Problem was that i used a power cord extension with a few other things on there like my monitor.

 

Whenever i would start a game. the gpu fans would turn on 100% (they are very loud) and i would loose monitor signal and basically crash, but still hear windows running apps in the background.

 

Plugged psu cable straight into the wall socket, no issues what so ever. Running on stock clocks, no crashes.

[BlackManINC]

You know, they could try not releasing these cards until they actually work out all the kinks. Releasing unfinished buggy games is already bad enough. Doing the same shit with processors is another story altogether. 

[iconfx]

I just made an account here to write what I experienced, in case anyone is interested.

Currently running 1080ti EVGA and already ordered RTX 3080 MSI Ventus 3X OC, I thought I would upgrade my Nvidia driver while I am waiting for the new card...and have experienced very similar and never experienced crashes on a same machine I have been running MW for the past 8 months.

PSU: 1200W 80+ Platinum model

CPU: Ryzen 1700

GPU: EVGA 1080ti Hybrid SC2

 

I just downgraded to older version of driver to see if the random crash is happening....

 

 

[Apington]

10 hours ago, BlackManINC said:

You know, they could try not releasing these cards until they actually work out all the kinks. Releasing unfinished buggy games is already bad enough. Doing the same shit with processors is another story altogether. 

Totaly with you on this one. Sending back my MSI Ventus OC today. Not gonna bother with a downclocked device I just bought for this kind of money. Definitely the last time I did the beta test for hardware this early.

 

Back to my 1070 ROG Strix I guess ¯\_(ツ)_/¯

[Marcelo oRieN Santos]

I had crashes in the start if ou check the page 1 and 2 there are videos also. I upgraded my PSU and the PC stoped shuting down.

Other than that i didn't have any crashes while in games.

What you guys can try.

 

• Clean Windows Install.
• Use the Studio Drivers
• Use auto fan curve so the GPU doesn't go above 80ºC

 

• I haven't tried Ray Tracing ... maybe try to shut off those options in games.
• Limit frame rates to the cap of your monitor with v-sync off.
• Disable Hardware-Accelerated GPU Scheduling

I dont have any more sugestions i have played for a while Far Cry 5 with clocks of 2050Mhz but the usage was about 60% ~ 70%

I played Borderlands 3 but the card didn't boost that high there.

 

[BlackManINC]

2 hours ago, Apington said:

Totaly with you on this one. Sending back my MSI Ventus OC today. Not gonna bother with a downclocked device I just bought for this kind of money. Definitely the last time I did the beta test for hardware this early.

 

Back to my 1070 ROG Strix I guess ¯\_(ツ)_/¯

And this is ironically the same kind of shit people always berated AMD for, releasing GPU's with bad drivers and such. I wonder what the excuse is gonna be now its Nvidia. 

[DarkvaleN]

 

On 9/26/2020 at 2:45 PM, BlackManINC said:

Totaly with you on this one. Sending back my MSI Ventus OC today. Not gonna bother with a downclocked device I just bought for this kind of money. Definitely the last time I did the beta test for hardware this early.

Same here. I sent my ventus back, since I'm fed up of wasting my time beta testing.  Will wait until stock goes back to normal and scalpers find something else to focus on and will buy hopefully a better tested RTX 3080 with more mature drivers.

[Abdullah Bhutta]

Please see this video for those who have RTX 3080 and its crashing under load and in games...

 

[1988fido]

On 9/29/2020 at 8:06 PM, Abdullah Bhutta said:

Please see this video for those who have RTX 3080 and its crashing under load and in games...

 

there is new video that correct that one.

it appears majority of early crashing's was fixed with nvidia driver update.

just small minority was from Aib bad manufacturing and the caps wasnt even the thing. in general I was taking some courses back in the day In Electrical and electronics stuff and did 3 years bachelors EEE but didnt finish due to some issues, so I got some idea. looked at schematics of cards, changed chips,caps,etc.. am not expert but was taking classes from expert and I know for a fact with my own eyes and hands that every small part on the board effect the card

so yeah the better caps = the better it is kind of ,.....but doesn't mean its 100% crashing it. it needs way in depth analysis than any youtuber did , heck even schematics for gpu's are a unicorn . you dont get them. unlike laptops can find many of them on russian sites.

so that method of using oscilloscope and tracing stuff with reverse engineering to truly find whats the issue not done yet , at least I looked into most of the youtubers i know.

[Abdullah Bhutta]

4 minutes ago, 1988fido said:

there is new video that correct that one.

it appears majority of early crashing's was fixed with nvidia driver update.

just small minority was from Aib bad manufacturing and the caps wasnt even the thing. in general I was taking some courses back in the day In Electrical and electronics stuff and did 3 years bachelors EEE but didnt finish due to some issues, so I got some idea. looked at schematics of cards, changed chips,caps,etc.. am not expert but was taking classes from expert and I know for a fact with my own eyes and hands that every small part on the board effect the card

so yeah the better caps = the better it is kind of ,.....but doesn't mean its 100% crashing it. it needs way in depth analysis than any youtuber did , heck even schematics for gpu's are a unicorn . you dont get them. unlike laptops can find many of them on russian sites.

so that method of using oscilloscope and tracing stuff with reverse engineering to truly find whats the issue not done yet , at least I looked into most of the youtubers i know.

This vid is outdated(i know)