What is considered a high system/cpu/gpu temp? What fps should it be locked at? Like in general.
Disabling NVIDIA Reflex Low Latency knocked 20c off my card temp, that setting seems to be responsible for a lot of overheating.
1 Like
I feel comfortable as long as you keep temp below 80c for both cpu/gpu. You should cap your fps at your monitor refresh rate. My monitor is 144hz so I just cap fps at 150.
playing the game right now, 3070 ti here, running 1440p, ultra, 45Âș Celsius 168.7w power draw, 40% fan speed
canât see how this game can push a card to die
your friend should look warranty and ask for a replacement cause itâs defective on his own fo sure
varies by gpu/cpu.
FPS shouldf be locked at whatever ur monitor is (if u can reach those).
gpu temps vary but most are designed for 83c being ânormalâ.
you obviously want em cool but as logn as not constantly 90c should be fine (though modern gpu are rated I believe 105c for shrt period of time)
cpu are similar. they can run at high temps (amdâs are designed to do so for performance reasons) but ppl dont liek it and try to keep em around 70-80c underload if ur on water.
I will take stuff that did not happen for 100$ alex.
Iâve mentioned this a few times in other threads, but in case you missed it, I recommend running MSI Afterburner with your NVidia GPUâs to better control their temps and power consumption. I currently down-volt my RTX2080ti OC to 85% of default power, 80c max temp, and always-on 60% fan max RPM (100% = hair drier noise levels). This is sufficient to keep the GPU fans relatively quiet, yet squeeze out 165fps @ 2k resolution on my current set-up. GPU temp maxes around 70c (usually mid 60âs), so itâs a good 10c below throttling temps.
I just gave a general advice that might keep your hardware alive bit longer when running hardware intensive software, but it will die eventually regardless.
Games or otherwise using a lot of ressources just point you toward the problem, they are not the primary cause.
I do wonder how many of the failing GPUâs had been used for Crypto-mining before becoming D4 work horses. Crypto-mining absolutely abused GPUâs 24/7 for many weeks or months at a time, then made their way into the secondary market after the big crypto-crash.
2 Likes
Maybe, but trying to find the original reason the hardware was weak is a waste of time.
If itâs new, they can contact manufacturer to change it. If itâs second hand, they accepted the risk this might happen in exchange for cheaper hardware.
Either way, it is not for Blizzard to save them. At best Blizzard could cap framerate to 120 by default, but the cards would still die down the line or on the next game.
Capping your framerate does not solve the issues this game causes as some are suggesting. There are serious bugs causing major problems. Even capping it in NVIDIA software does nothing to stop the heat ramp some people experience when doing things like opening their inventory or talking to certain NPCs.
I have this issue myself to a degree. My GPU will jump 10-20 degrees from certain UI elements being open even for less than a minute. Not combat. Not anything that should actually be doing anything at all. Certain UI elements stress the GPU more than combat in the densest dungeon while just standing in town. My in-game frame rate is capped at 100 currently, in NVIDIA control panel itâs capped higher at my monitors refresh rate. It shouldnât even have to be that low but I was trying to keep the heat from jumping so high for no reason. The heat isnât a problem until you talk to certain NPCs or for some people open the inventory.
Someone mentioned above NVIDIA reflex being a possible cause and I think that may be a factor. I have it on atm and I think I turned it off at one point, but my settings were reset either after the recent patch or updating my driver. After randomly turning various settings down I had significantly reduced the problem for awhile (temp would still spike from talking to certain NPCs, but not as much as before) and it came back recently so I think that may have been it.
And yes, we can say the actual hardware failures arenât directly caused by the game, but we can simultaneously acknowledge that this game is an unoptimized buggy POS thatâs causing stress on hardware that it shouldnât be and thatâs exasperating existing hardware problems for some people.
You shouldnât have to âCheck your temperaturesâ when you play a game.
Do you do this on console?
Computers and video games arenât made to âCheck your temperatures when usedâ
It doensât work like that.
It is good to check them once and a while to ensure your stuff is clean but we shouldnât have to do that.
Have you tried my MSI Afterburner suggestion? The recommended steps drop the GPUâs voltage while setting the fan speeds to always-on instead of periodic / reactive acceleration.
It does and it doesnt. It depends on how you wanted your equipment last. Also, in todays technology, GPUâs should have a safety net to avoid this kind of issue. But one thing I am sure of, technology specially machines like computers is like a LOTTERY, you either get the best or the worst one.
You dont overclock in console so you dont.
Interesting that there was this other game called New World, and people there were talking about that game frying their cards.
Now we are talking about Diablo 4 frying cards.
Itâs clear the common denominator is the cards, not the gamesâŠ
1 Like
I actually followed a guide on undervolting my GPU in aferburner to just keep the temps lower and while I think it worked most of the time, it doesnât stop the vendor/UI bug. Also tried setting up a custom fan curve and I guess that lowered the average temperatures the rest of the time, but it doesnât really prevent the vendor/UI issue. That drives my GPU temps up to 80c even if I undervolt it. I have no idea what itâs doing or how but thereâs something weird going on with certain UI elements, possibly only when combined with certain other settings and factors, but itâs pretty weird. Thankfully it doesnât kill my GPU or hasnât yet, but I find it abnormal/concerning. Based on those tech support threads it also seems to affect a decent chunk of people, causing more serious problems for some.
When I play later Iâll turn reflex off again and see if that stops it from happening.
Iâm do not recall seeing the same thing happen on my RTX2080ti OC at the vendor UI, but I will validate once I get home.
I would lock at 5 fps higher than your monitor refresh rate. Most are 60 or 120. Thereâs really no reason to go to a higher refresh than your display can support.
I think maybe some VR takes 240 (?) but I donât VR.
Same, with a PNY built RTX 2070S card.
You guys who dont recognize as a hardware failure crack me up