So I think my Vega64 is completely stuffed?

[Goose]

I have a reference Vega64 that I got second hand two or three years ago and I think it's probably come to the end of its life. I hadn't gamed in the last 3 months or so and when I decided to start again the other day with No Mans Sky and Hitman 2 both games would cause my system to reboot within 2 minutes or so following a very brief windows blue screen error message. I've never had crashes like this before, normally I would only ever get crashes to desktop as a result of an unstable overclock.

I resorted to doing a fresh format and reinstall of Windows 10 but the problem persisted. I've tried a few other games with the same result.

I've just disabled the Vega 64 and installed an old GTX970 from my dad's system and everything is working just fine which I think makes it fairly conclusive that my Vega64 is kaput?

It's not a heat issue either as I have it watercooled and I dont think its my PSU either as it's an 850W EVGA unit (albeit a 7 year old one).

If anyone could confirm (or not) my fears i'd really appreciate it.

 

[LtMatt]

Could be kaput.

You could try putting it in the oven for 30 minutes at 200c since the warranty has long since passed and you have nothing to lose regardless.

 

[eeii]

Even though your system is water called still check your temps as I had a water cooled Vega 64 which dried the thermal paste on the water block and had the same problem.
Use MSI after burner or the AMD driver overlay to check. I would also make sure you are using seperate GPU power cables. Try swapping the GPU power cables to different power supply sockets. Its worth a try.

 

[Aretak]

BSoDs with no other indicator are a fairly rare failure mode for a GPU. Since you have another system available, the easiest way to do things would be just put it in the other system and see if it exhibits the same behaviour. Sure it's a pain if it's water cooled, but you're going to be taking it out anyway if it's dead.

 

[pagepack]

As mentioned try the heat method if you have nothing to lose. I couldn't boot to desktop with my 290x, tried another gpu, everything was fine. Used a heat gun as per various YouTube videos on my card and everything is working now.

I'm sure it's a temporary fix, hopefully until i can grab an mrsp gpu....
..

 

[Goose]

Even though your system is water called still check your temps as I had a water cooled Vega 64 which dried the thermal paste on the water block and had the same problem.
Use MSI after burner or the AMD driver overlay to check. I would also make sure you are using seperate GPU power cables. Try swapping the GPU power cables to different power supply sockets. Its worth a try.

Temps were always in the 40s so no problems there and i've always been using separate cables for each power connector.

BSoDs with no other indicator are a fairly rare failure mode for a GPU. Since you have another system available, the easiest way to do things would be just put it in the other system and see if it exhibits the same behaviour. Sure it's a pain if it's water cooled, but you're going to be taking it out anyway if it's dead.

Maybe i'll give that a go. I should have said in my OP that the card was formerly in a mining rig and I was never able to overclock it even by a tiny bit as it would artefact and/or crash. I could undervolt it a bit on stock clock speeds but even on stock volts trying to overclock it even 1mhz would make it unstable. It never bothered me too much having to run it at stock speeds since the card was super cheap owing to its usage in a mining rig. Basically, knowing its history and my experience with it i'm fairly confident the card is now a goner.

As mentioned try the heat method if you have nothing to lose. I couldn't boot to desktop with my 290x, tried another gpu, everything was fine. Used a heat gun as per various YouTube videos on my card and everything is working now.

I'm sure it's a temporary fix, hopefully until i can grab an mrsp gpu....
..

Is buying a card at MSRP even possible any more?

 

[Aretak]

Maybe i'll give that a go. I should have said in my OP that the card was formerly in a mining rig and I was never able to overclock it even by a tiny bit as it would artefact and/or crash. I could undervolt it a bit on stock clock speeds but even on stock volts trying to overclock it even 1mhz would make it unstable. It never bothered me too much having to run it at stock speeds since the card was super cheap owing to its usage in a mining rig. Basically, knowing its history and my experience with it i'm fairly confident the card is now a goner.

Given that experience, have you tried underclocking it a bit to see if that makes any difference?

 

[Goose]

Given that experience, have you tried underclocking it a bit to see if that makes any difference?

Yep

 

[pooley]

Yep

your card does sound goosed me old duck