4090 Woes!

[CrazyArsedMonky]

Greetings folks!

I only ever post if i am in the process of upgrading, or when i have a problem - unfortunately, this post is the latter...

My Asus TUF 4090 has started playing up - crashing when under stress, in various games.

It seemed to coincide with a driver push, so I rolled back to an older driver and crossed my digits - unfortunatley it still crashes.

The failure mechanic seems pretty common based on some Googlin' - black screen, monitor toggles between connected / unconnected and the PC is unresponsive (i have to hold reset or power switch).

Edit - The GPU fans never go crazy, but did once post crash. After a post-crash restart the BIOS didn't recognise the GPU. To recover, i turned computer off, waited a minute, then turned on again and it worked fine.

To top it all off, the card is 10 days out of warranty (yup...), and i have been told to go to Asus with my begging bowl, but there is no number to phone - just a digital postbox, so would no doubt wait a week or two for them to say 'suck it up'.

I was snooping around the forum for similar occurances, and saw that someone had replaced the thermal paste (not for this fault) - which seems like a pretty darned good idea, just wondering if maybe it is thermal induced crash that i might be able to fix...

When i open the case there isn't a smell (indicating a popped cap or such like), so it might not be toasted, but i wanted to get other opinions before i start open heart surgery.

As always, advice and feedback greatly received!

Thanks in advance

 

[Kelt]

Is your PSU up to scratch?

Black screening is a common symptom caused by a failing/not supplying enough power, PSU.

 

[Livewire88]

If it were me, I would exhaust trying to claim under warranty first which I guess will be unsuccessful. If so open the card, repaste and cross your fingers it fixes the problem.

 

[CrazyArsedMonky]

Is your PSU up to scratch?

Black screening is a common symptom caused by a failing/not supplying enough power, PSU.

Thanks for the reply Kelt! I think so, it is reasonably old, but has been working A-ok for the past few years (fault started about 2 weeks ago), had the 4090 3 years.
Is there a way i can test the PSU outputs?

Another few nuggets i missed:

The GPU fans never go crazy, but did once post crash.
After a post-crash restart the BIOS didn't recognise the GPU. To recover, i turned computer off, waited a minute, then turned on again and it worked fine.

 

[Tetras]

My Asus TUF 4090 has started playing up - crashing when under stress, in various games.

I was snooping around the forum for similar occurances, and saw that someone had replaced the thermal paste (not for this fault) - which seems like a pretty darned good idea, just wondering if maybe it is thermal induced crash that i might be able to fix...

Have you been looking at the temps with something like hwinfo? Usually you'll find that if this is a common problem a google of the card model will suggest other people experiencing it and what the main symptom is (e.g. high hotspot).

Is there a way i can test the PSU outputs?

You can check the software readings, but they're not very reliable.

One way would be to downclock/power limit the card heavily and see if that stops the crashes, though that's not exactly fool proof because you're also understressing the card.

Have you checked the connector (power) is in good condition?

 

[JonRGV250]

Have you checked the connector (power) is in good condition?

This ^
I had a nightmare with my Zotac 4090 causing all kinds of weirdness, until I swapped out the crappy CableMods power cable, and after that it was as good as gold.

 

[GeorgeStorm]

What do the temperatures show when under use? Or any change in clocks speeds etc?
Have you tried downclocking to see if it has any impact?

 

[CrazyArsedMonky]

@Tetras Great call, I just whipped the cover off - all connectors are are in good condition at both ends, PSU (Corsair 850 jobber) is fine from outside and no whiffy smells.

@GeorgeStorm George - you beat me too it! I was just doing some testing with Sins of a Solar Empire II - running in window mode. Temps were around 55C with lots going on. I got several error messages, then a period of plain sailing - then crash. I was Alt-tabbed writing this reply at the time, so didn't see temp when it crashed, but case was cool, cards and cpu were fine - and the fans were pretty relaxed...

GPU is currently in stock mode, so no overclocking at all (I never overclocked it, had no need to). I haven't tried underclocking it - will investigate that...

Thanks all!

 

[Purgatory]

To top it all off, the card is 10 days out of warranty (yup...), and i have been told to go to Asus with my begging bowl, but there is no number to phone - just a digital postbox, so would no doubt wait a week or two for them to say 'suck it up'.

ASUS Website Information - Contact ASUS

ASUS is one of Fortune magazine’s World’s Most Admired Companies, and is dedicated to creating products for today and tomorrow’s smart life. Our comprehensive portfolio includes Zenbo, ZenFone, ZenBook and a range of IT devices and components, along with AR, VR and IoT. ASUS employs more than...

www.asus.com

Just going to leave this here if anyone else is confused how to contact ASUS in UK.

For queries related to the ASUS eShop, please contact customer service via email at [email protected], or call 0208 551 6505.

Mon-Fri 9am to 5pm (Excluding Bank Holidays)

For general queries and support, please contact 0800 0260 440 or 01442 265548, or visit www.asus.com/uk/support/.

 

[BigBANGtheory]

the 1st thing to do is blast some air down the cooling fins to clear out dust/debris/pet hair etc