AMD Crimson Software Might Cause locked fan RPM at 20% (overheats cards)
Click here to post a comment for AMD Crimson Software Might Cause locked fan RPM at 20% (overheats cards) on our message forum
Darkest
The majority of people who play games on the PC are far removed from the enthusiasts you find on this forum. They just start up games and play them, they don't do anything even remotely technical, at most they'll update their drivers on a regular basis. Spend five minutes checking random game forums on Steam, you'll see exactly what I mean. I'd actually wager that most don't even know how to monitor temperatures.
(.)(.)
Yeah thats im seeing, the steam forum level of user competence.
---TK---
They beta test these drivers out or what? That is a huge freaking bug there. Nvidia had a driver out like that years ago but with me monitoring temperatures all the time risk is minimal. Though I was not affected by that driver.
IcE
To be fair I don't monitor temps anymore. The card has protections anyway, it will lock up once it hits 100C to prevent damage.
Only way I could see any cards dying would be to a bug like this existing for a really long time, which obviously isn't the case here.
Redemption80
Surely the cards would throttle to avoid damage?
Either some people have noticed this bug and decided to try and see what they can get out of it, or there is something else in the drivers stopping the card throttling.
sverek
Looks like someone at AMD getting "fired"...
[spoiler]failed with puts on glasses meme...[/spoiler]
Clouseau
They also need to alter the disclaimer that is agreed to before enabling overdrive. Currently they only mention that the cpu may be damaged when operating outside of stated specifications. Legal team feel asleep on that one.
sykozis
The "average user" isn't going to immediately relate an apparent GPU failure to a driver. It would take someone technical minded to relate the 2. The people most likely to be affected by this, are the least likely to enable OverDrive and anyone dumb enough not to monitor their GPU temp after enabling an overclocking feature deserves to have their GPU fail imho....
When I overclock, I monitor temps at every adjustment. WTF overclocks purely by adjusting clock speeds? A lot of the cards on the market will hit a thermal wall before they reach the limit of the GPU thanks to piss poor reference coolers.
PhazeDelta1
This just validates my point about AMD and their driver team even more. What a bunch of knobs.
Rugburn
Although I appreciate the protections in place on modern gpus. I still make it a habit to check temps, vram and gpu usage while playing.
Stormyandcold
The only thing with drivers that we should fear...heat-of-death. Both sides have been guilty of it now, but, it's still a major over-sight especially with all the hype. Someone didn't do the checklist it seems. This is why they need to be more open and have wider beta testing teams so as to avoid public backlash when things don't work. Potential critical failure of hardware is serious business.
Neo Cyrus
I don't understand how such a thing could get through, how could it happen... this isn't the first time either. It's happened in the past with both nVidia and AMD.
Dch48
I'm not having any problem. My fan speeds and temps are fine. I don't use Overdrive though and never overclock anything.
Fox2232
My HD7970 did lock up on 91°C for GPU or if VRMs hit 95°C 1st.
I am sure it was not driver level lockup overridable by driver. So while this driver keeps fans at low spin, it will not fry GPU.
My Fury X fan stays reportedly at 20% but even then GPU temperature stays at around 42~45°C in gaming (PoE, KF2).
We have here user, I think "Agonist", who claims that it killed his 2 GPUs (liquid cooled).
I really wonder how did AMD's crimson driver take over control of pump and fans for that water cooling he used.
And Even more I wonder if it did actually boil before failing, because temperatures required to permanently damage GPU are above boiling point of water which is used in LCS.
Running Afterburner's custom fan profile works as usual.
fantaskarsef
This only makes me wonder since when AMDs drivers suffered from their teams working on mantle / vulkan / dx12.
Good thing there'll be a hotfix for this and it doesn't come all too late, but if I'd have an AMD card, I'd still be annoyed by this.
Fox2232
At least it was not WHQL, or was it? I think people who has been hit by this issue are people who may rethink their beta driver usage.
And as I wrote above, your card can lockup itself due to temperature trigger 1000 times before frying. Yet people claim 1 lockup and dead?
I think some people trying to surf on this issue and get replacement for their already dead cards. Others are looking for issue source where it did not came from. (Instead on checking their own handiwork.)
fantaskarsef
Yeah, sounds pretty logical. Still, stories such as this tend to make people worry even if there's not too much reason to do so (especially those not as enthusiastic about our hardware and everything that comes with it).
Passus
just noticed
mine locked itself to 60% when i have max 45% on any profile
it had locked itself in the global overdrive settings, this is a fresh install of the driver after windows 10 went barmy yesterday and i had to reset it
also it had locked the power to +20% even though i had not touched the overdrive prior to noticing this morning
RealNC
Now I know why they named the new software "Crimson". It's the color of the glow the GPU gives off after installing the driver. :bolt:
joe187
AMD Crispy, cause some just like it HOT!
:leave:
