FTW3 Ultra 3090 constant crashing!

shadow2761

Limp Gawd
Joined
Mar 11, 2016
Messages
455
Just recieved this card yesterday and am very upset already because games are constantly crashing to desktop with atleast one BSOD aswell so far. My temps seem to be fine @ 75°C max under load and clock speeds are out of the box around 1800-1900 MHz tops.

I have the latest 457.09 drivers and I thought the crashing shenanigans have been addressed weeks ago with driver updates.

But my card is crashing like no tommorow in PUBG and Remenant from The ashes game, where as the MSI 2080 Ti I had just prior to installing the 3090 was running crash-free for almost 2 yrs.

So is anyone else experiencing tons of crashing like this still?

Or do I have a dud card and need to RMA?
Is it possible I did not sit the card fully into the PCIe slot because when I was installing this card into the PCIe slot, it felt like the left side of the GPU PCIe connection pins did not fully seat down into the slot. I applied a fair amount of pressure to make sure it would go all the way in.

What do you guys rekon?

EDIT: SOLUTION for my Case is below.
 
Last edited:
  • Like
Reactions: erek
like this
What kind of PSU do you have? If it does it during when ever you put a load on it makes the PSU suspect.
 
Power supply is a big thing. In my testing you will need to ensure the 3 pcie power ports are tied to at least 40a per plug.

Pcie1 pulls 120 to 135w, pcie2 pulls 120 to 135w, pcie3 pulls as much as 160w.

A friend with a 1kw psu noted of 4 pcie plugs he had only 2 were 40a and 2 were 32a. Refer to your psu manufacturer for how many amps you have available per port.

It also looked liked my ftw3 came with xoc bios out of the box because in my initial superposition runs I pulled 475w peak. I later pulled 481w playing ultra mode watch dogs legion during normal gameplay.

To add insult to this, there is barely a 400pt score difference in superposition between a ftw3 and xc3 ultra at stock speeds. The xc3 will run 360w all day at 79c and give 99% of the perf of the ftw3 when gaming.
20201103_125355.jpg20201031_005203.jpg
 
Last edited:
use gpuz to make a log file (sensors tab -> log to file at bottom) and post it here of the crash.
 
  • Like
Reactions: erek
like this
Thanks for replies guys.

I have a EVGA SuperNova P2 1000W PSU.

Does anyone have any information how much amps each port has for this PSU?
 
I had the same PSU before my 1200 P2 upgrade a few years ago. All power ports for VGA have access to the same rail and it has 999W available on all those plugs though for wire rating safety, I would ensure no splitter and each 8pin gets its own dedicated plug.

The funny part is my manual for the P2 1200 IS the manual for the 1000 as well.
 
Thanks for replies guys.

I have a EVGA SuperNova P2 1000W PSU.

Does anyone have any information how much amps each port has for this PSU?

That's an excellent psu. Troubleshoot time, do you have another rig to test said card in?
 
If your CPU is overclocked, try running it stock. If its a PSU problem, that might ease the load enough of figure it out. Maybe even disable turbo, for testing purposes.
 
If your CPU is overclocked, try running it stock. If its a PSU problem, that might ease the load enough of figure it out. Maybe even disable turbo, for testing purposes.
CPU is not overclocked for this purpose, stock speeds. I think I have DDR4 ram @ 3600MHz XMP profile though, should I disable it?


That's an excellent psu. Troubleshoot time, do you have another rig to test said card in?

I am going to take the card to my friends computer over the weekend. He just recieved his STRIX OC 3090 and did not report any issue with it so far. So I guess I can get him to remove his 3090 and pop mine directly in and test out some games.
 
Well you could reduce your power target to see if its PSU related. Just run OC Scanner on MSI Afterburner and set the power target to 80%. You won't lose any performance from stock.
 
Well you could reduce your power target to see if its PSU related. Just run OC Scanner on MSI Afterburner and set the power target to 80%. You won't lose any performance from stock.
Would this be because of a possible faulty PSU, or is the P2 1000w not enough?

Anyways, I guess I will try this aswell.
 
Back when I got my original P2 1000, there were some reports of badly made cables for it. Some people reported that there was arcing going on. My P2 1000 died years ago and when it was replaced by EVGA, they sent me a 1200 P2 as replacement. I never questioned the upgrade and that 1200 has been good for me the past 4 years now carrying me from 980, 1080, 1080ti, 2080ti, 2080ti sli and 3090.

Do some inspection of the cables for signs of any arcing or shorts.
 
Reseat the card, some people have run into this issue when the cards pcie pins weren’t fully making contact with the slot.
 
Just recieved this card yesterday and am very upset already because games are constantly crashing to desktop with atleast one BSOD aswell so far. My temps seem to be fine @ 75°C max under load and clock speeds are out of the box around 1800-1900 MHz tops.

I have the latest 457.09 drivers and I thought the crashing shenanigans have been addressed weeks ago with driver updates.

But my card is crashing like no tommorow in PUBG and Remenant from The ashes game, where as the MSI 2080 Ti I had just prior to installing the 3090 was running crash-free for almost 2 yrs.

So is anyone else experiencing tons of crashing like this still?

Or do I have a dud card and need to RMA?
Is it possible I did not sit the card fully into the PCIe slot because when I was installing this card into the PCIe slot, it felt like the left side of the GPU PCIe connection pins did not fully seat down into the slot. I applied a fair amount of pressure to make sure it would go all the way in.

What do you guys rekon?
Im crashing in GTA V on a Gigabyte Vision 3090. I also have direct confirmation of another owner experiencing the crashing

This seems close to class action terroritory
 
Reseat the card, some people have run into this issue when the cards pcie pins weren’t fully making contact with the slot.

This is interesting. Because when I installed this huge card, I could not bloody tell how well the pcie pins were making contact with the slot.

To me it looked like the left side of the GPU was slightly not fully in so I tried to apply alot of pressure on it to make sure it is fully seated in. But still I have no idea if it is because the card and its physical components are so big!

Are you able to link me to other peoples post who was having this problem until they seated the card in properly?
 
This is interesting. Because when I installed this huge card, I could not bloody tell how well the pcie pins were making contact with the slot.

To me it looked like the left side of the GPU was slightly not fully in so I tried to apply alot of pressure on it to make sure it is fully seated in. But still I have no idea if it is because the card and its physical components are so big!

Are you able to link me to other peoples post who was having this problem until they seated the card in properly?

I read it somewhere on Reddit and since i scroll through tons of posts daily there’s no way I’d be able to find it. Give it a shot and see if it works.
 
Today I have

-tried the evga Normal and OC Bios updates for FTW3 Ultra 3090.
-checked event viewer, nothing for the time of each crash
-reverted to previous 456.71 drivers
-underclocked the GPU by 100MHz

Nothing is working. This card is constantly crashing every 5-15minutes of gaming.

Will give this card to a friend to try on his machine on Monday.

:(
 
  • Like
Reactions: erek
like this
Im crashing in GTA V on a Gigabyte Vision 3090. I also have direct confirmation of another owner experiencing the crashing

This seems close to class action terroritory
Yeah, no.

It's getting close to RMA time.
 
like I said ^
Sorry, I tried to log the sensors but I accidentally started logging them after the crash lol

I thought it was going to save all the data already done when i clicked the log to file, but no it starts the logging from only when you click it. So I will log it properly tommorow.

And i will also update ASUS Z390 Strix E motherboard BIOS, I haven't updated the MB in over 8 months.

And I noticed my card is using a board power draw of 430W according to GPU-Z under load and I haven't increased the factory OC setting. Is this normal?
 
Sorry, I tried to log the sensors but I accidentally started logging them after the crash lol

I thought it was going to save all the data already done when i clicked the log to file, but no it starts the logging from only when you click it. So I will log it properly tommorow.

And i will also update ASUS Z390 Strix E motherboard BIOS, I haven't updated the MB in over 8 months.

And I noticed my card is using a board power draw of 430W according to GPU-Z under load and I haven't increased the factory OC setting. Is this normal?
click the reset button in afterburner to reset it to stock again. It's possible it's autoloading a profile on startup. But get a gpuz log and I'll look at what's happening.
 
The upper class 3090 cards will pull 400-440W in stock forms while the lower tier ones will only draw around 340-360W.

All tiers of 3090 seem to be able to boost to 1900-2000mhz when left at stock from my experience. (Asus TUF, EVGA XC3, FTW3)

Seating the 3090 FTW3 is a bit difficult. In my case, I am using a vertical mount and the locking clip seems to not be latched in 100%, but I have not encountered any stability issues. The sheer size and weight make it difficult to get a feel for how well it is plugging into the slot.
20201106_100655.jpg
 
Last edited:
Today I have

-tried the evga Normal and OC Bios updates for FTW3 Ultra 3090.
-checked event viewer, nothing for the time of each crash
-reverted to previous 456.71 drivers
-underclocked the GPU by 100MHz

Nothing is working. This card is constantly crashing every 5-15minutes of gaming.

Will give this card to a friend to try on his machine on Monday.

:(
any update?
 
click the reset button in afterburner to reset it to stock again. It's possible it's autoloading a profile on startup. But get a gpuz log and I'll look at what's happening.
Thank you sir, appreciate it. Will get a GPUz log out today hopefully.

The upper class 3090 cards will pull 400-440W in stock forms while the lower tier ones will only draw around 340-360W.

All tiers of 3090 seem to be able to boost to 1900-2000mhz when left at stock from my experience. (Asus TUF, EVGA XC3, FTW3)

Seating the 3090 FTW3 is a bit difficult. In my case, I am using a vertical mount and the locking clip seems to not be latched in 100%, but I have not encountered any stability issues. The sheer size and weight make it difficult to get a feel for how well it is plugging into the slot.View attachment 296472
Ok thanks for the info about the power draw.
Yes I also can't tell if I have seated the card in properly because its huge. I might try another PCIe slot today aswell, the other 16x slot in the Z390 STRIX E motherboard (PCIE 3.0 X16_2)

any update?
Nothing new yet, gonna try some things later today hopefully as mentioned above.
 
This may have nothing to do with your issue, but I installed my 3090 FTW3 it worked perfectly fine on shipped bios without Precision X1. I installed the beta bios and the newest Precision X1 and I've had numerous soft crashes ever since. Screen would go black but eventually come back after a few seconds. Very annoying. This never happened gaming, only when doing normal stuff on the computer like surfing the internet and youtube. I uninstalled Precision X1 and reverted back to the shipped BIOS and have had zero issues in several hours.

This exact same behavior happened with my 3080 FTW3 in this same computer. Something I have does not like Precision X1 at all. I'm done with it.
 
Today I have

-tried the evga Normal and OC Bios updates for FTW3 Ultra 3090.
-checked event viewer, nothing for the time of each crash
-reverted to previous 456.71 drivers
-underclocked the GPU by 100MHz

Nothing is working. This card is constantly crashing every 5-15minutes of gaming.

Will give this card to a friend to try on his machine on Monday.

:(
when's the last time you updated your motherboard bios?

*I just saw that you are going to do that. Good. Maybe there is a compatibility issue which was fixed. Ampere is PCI-E 4.0, afterall.
 
*PROBLEM FIXED* :)

Ok guys, can confirm 100% fixed now, just did some heavy gaming for about 5 hrs on many games, 0 crashes!
Boosted to 1950MHz, temps around 65-72°C under load.

Solution was simple, it was either one of these two thing I did, not sure which one was the actual fix:

1. Updated motherboard BIOS. Mine was over 8 months out of date. Not sure how I let that one slip, i usually keep everything up-to-date.

2. Updated windows version to 20H2 via Update assistant because the automatic windows update did not go up to this version.

Thanks to everyone helping out here, I am so glad I don't have to RMA the card!
 
*PROBLEM FIXED* :)

Ok guys, can confirm 100% fixed now, just did some heavy gaming for about 5 hrs on many games, 0 crashes!
Boosted to 1950MHz, temps around 65-72°C under load.

Solution was simple, it was either one of these two thing I did, not sure which one was the actual fix:

1. Updated motherboard BIOS. Mine was over 8 months out of date. Not sure how I let that one slip, i usually keep everything up-to-date.

2. Updated windows version to 20H2 via Update assistant because the automatic windows update did not go up to this version.

Thanks to everyone helping out here, I am so glad I don't have to RMA the card!
#Jealous

i'm still troubleshooting my own instability
 
*PROBLEM FIXED* :)

Ok guys, can confirm 100% fixed now, just did some heavy gaming for about 5 hrs on many games, 0 crashes!
Boosted to 1950MHz, temps around 65-72°C under load.

Solution was simple, it was either one of these two thing I did, not sure which one was the actual fix:

1. Updated motherboard BIOS. Mine was over 8 months out of date. Not sure how I let that one slip, i usually keep everything up-to-date.

2. Updated windows version to 20H2 via Update assistant because the automatic windows update did not go up to this version.

Thanks to everyone helping out here, I am so glad I don't have to RMA the card!
Very likely it was the MB BIOS update, as I've seen a few other people fix similar stability issues with GPUs by doing that.

Glad you got it squared away.
 
Strange, what card do you have?

And you have tried updating everything? Motherboard BIOS, Windows etc?
with the gigabyte cards, check the pins on the power connectors. Been some issues with how they did the connectors that the pins not engaging 100%.

See this thread to understand what you are looking for.
https://www.overclockers.co.uk/foru...-gaming-oc-power-connector-concerns.18900176/
GeForce Gigabyte RTX 3080 VISION OC This card is not effected by the issue
GV-N3080VISION OC-10GD

Mine's a 3090, 2042, so WK 42?


shadow2761 did you install this as part of resolving your issues? 2020-10 Cumulative Update Preview for Windows 10 Version 20H2 for x64-based Systems (KB4580364)
 
#Jealous

i'm still troubleshooting my own instability
Strange, what card do you have?

And you have tried updating everything? Motherboard BIOS, Windows
GeForce Gigabyte RTX 3080 VISION OC This card is not effected by the issue
GV-N3080VISION OC-10GD

Mine's a 3090, 2042, so WK 42?


shadow2761 did you install this as part of resolving your issues? 2020-10 Cumulative Update Preview for Windows 10 Version 20H2 for x64-based Systems (KB4580364)

I have to check if that was installed.
But I did use the windows update aswell the update assistance tool to get as many updates as possible, so yea that possibly could have been aquired.

I will have to check later on.
 
Strange, what card do you have?

And you have tried updating everything? Motherboard BIOS, Windows


I have to check if that was installed.
But I did use the windows update aswell the update assistance tool to get as many updates as possible, so yea that possibly could have been aquired.

I will have to check later on.
Gigabyte VISION 3090

I'm scared to update my BIOS due to instablity:


https://hardforum.com/threads/asus-...v2-pi-1-1-0-0-patch-c.2003188/post-1044798079
 
Back
Top