• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

576.02 Nvidia Drivers

RTX 3090 here. With 572.83 I was getting random black screens in Atomic Heart, almost as if my GPU was overheating because the image would re-appear, but it didn't affect Ghost of Tsushima that I played for 40+ hours. I rolled back to 566.36(?) and Atomic Heart ran fine. Decided to update to 576.02 and Atomic Heart is doing well with no random black screens.

I think that having a DisplayPort and HDMI display connected was a condition that created a black screen, that's what I have. I won't be surprised if the issue returns.
 
The list of fixes includes: "RTX 50 series displays blank screens on LG 5k2k 45GX950A-B when running in DisplayPort 2.1 mode w/ HDR". I just tried it, and the problem is still there.
 
Bigmonitorguy said:
The list of fixes includes: "RTX 50 series displays blank screens on LG 5k2k 45GX950A-B when running in DisplayPort 2.1 mode w/ HDR". I just tried it, and the problem is still there.
Click to expand...
I can confirm that this new Nvidia driver didn't fix this issue. Switch to the supplied HDMI cable and you should be good.
 
TaintedSquirrel said:
Seeing reports of big performance gains on these.

https://www.reddit.com/r/overclocking/comments/1k0pbbj/just_hit_10k_in_steel_nomad_after_the_latest/
Click to expand...
That comes at a cost it seems. Unwinder (creator of MSI Afterburner) posted about this on reddit:

Not true. It is by far not "only" Afterburner as you wrongly claim. NVIDIA's official NvAPI_GPU_GetThermalSettings API is broken in this driver, which affects thermal monitoring in all tools using it (which includes OS task manager, PresentMon and every single hardware vendor's specific software like ASUS GPU tweak, EVGA Precision X1 or MSI AB).
NVIDIA also has undocumented private thermal monitoring API, NvAPI_GPU_ThermChannelGetInfo (which includes hotspot and VRAM monitoring interface), which still works in this driver and which can be optionally used by GPUZ or HwInfo, but ironically it is reverse engineered unoffficial fuctionality and NVIDIA doesn't allow partner's software to rely on it (that's exactly why you never see VRAM temperature in any partner's software, but can see it in independent ISV's products like HwInfo).
Also, stuck thermal sensor is just a part of problem. When the sensor is stuck in this driver, it also blocks driver's thermal GPU boost V/F curve adjustment. So GPU starts boosting much higher than before (due to not seeing temperature increase). So it bumps performance but may negatively affect stability of some PCs.
Click to expand...
https://www.reddit.com/r/nvidia/com...utm_medium=web3x&utm_name=web3xcss&utm_term=1

Stuck thermal sensor is just a part of problem. When it stops reporting data, NVIDIA driver's GPU boost thermal adjustment for V/F curve stops working too, causing GPU to clock higher than it did before. I summarized it here:
Click to expand...
https://forums.guru3d.com/threads/n...l-driver-download.456063/page-12#post-6326425

I am staying clear of those drivers, since it would be on a "4070" laptop I also use for work. Dont want to cook it. :p
 
Tamlin_WSGF said:
That comes at a cost it seems. Unwinder (creator of MSI Afterburner) posted about this on reddit:


https://www.reddit.com/r/nvidia/com...utm_medium=web3x&utm_name=web3xcss&utm_term=1


https://forums.guru3d.com/threads/n...l-driver-download.456063/page-12#post-6326425

I am staying clear of those drivers, since it would be on a "4070" laptop I also use for work. Dont want to cook it. :p
Click to expand...
Verified this as well. Afterburner shows 30c no matter what you're doing. GPUZ is showing the correct temps.
 
TaintedSquirrel said:
Great now people are going to whine about Nvidia gimping performance when they fix this.
I assume a hotfix is coming in the next few days. Broken temp sensors can't be ignored.
Click to expand...
I hope so. Games work fine on older drivers, but would be fun to play around a bit with DLSS 4, so I am waiting for a stable driver that supports it. People have success unsticking thermal sensor by reinstalling AB over the existing, but who knows when it gets stuck again. According to Unwinder, its not an error in AB, but drivers, so its not a fix.
 
Any issues on RTX 4*** series with these drivers? 572.70 works great for me after some issues with the prior release.
 
RTX 3080 Ti FE user here. Been getting fairly consistent crashing on every single game I have played on my PC since my last driver update. I was playing just one game for a while so thought it was a game issue, but once I started playing others and they were exhibiting the same problem, I am now suspecting drivers.

Don't know for sure at this point, but rolled back to 566.36 last night, so will see if I have the same problems. If not, then going to blame the newer drivers not playing nice with Windows 11 24H2. If I still have problems then may look at just reinstalling Windows at this point. But been horrible trying to game the past 2-3 weeks:

  • Back4Blood
  • Witcher 3
  • Fallout 76
  • Dungeon Alchemist
  • several smaller indie titles
All crashing.

Card itself is fine, temps are good (custom water loop), verified nothing else odd going on hardware wise. Removed any custom tweaks on RAM and went back to DOCP/XMP, put CPU back on stock with auto PBO settings, and put the 3080 Ti back to stock settings with no PL increase with my normal +120/+1000 overclock. As for Windows, checked updates, virus scanner, etc. So was at a complete loss to my sudden instability.
 
Last edited:
Is there a way to intentionally block the thermal sensor?
Disabling the driver's built-in throttling could help stabilize OC's.

I am being... half serious.
 
Flogger23m said:
Any issues on RTX 4*** series with these drivers? 572.70 works great for me after some issues with the prior release.
Click to expand...
I'm sure other folks are having issues. None yet for me with the GPU sensors nor with any games - this is the second 50 series driver that I was able to install without needing to DDU first. I did decide to try out the studio driver this time.
 
sk3tch said:
YES!
Click to expand...
The Fortnite fix is good - I had one night (Wednesday) with zero crashes. Then last night I had a crash towards the end of my time to game...just took me out of it again. So annoying. RTX 5090 Astral - I think I am going to return it to MC I am still in my window. Played with an LLM for a bit which was cool but of course it cranked the amps to the 12VHPWR pins.

I got a 5070 Ti from another member (was gonna upgrade my son's rig, heh) going to see how I can "live" with that for the last few days of the return window for the 5090. If it is totally fine (which it will be) - punt the too expensive GPU.
 
“NVIDIA's Latest 576.02 WHQL Driver Add Up to 8% Performance Bump in Synthetic Benchmarks
by AleksandarK Today, 04:21 Discuss (11 Comments)
NVIDIA's latest GeForce 576.02 WHQL driver, released on April 16, which brought support for the GeForce RTX 5060 Ti, appears to deliver unexpected performance gains for several "Blackwell" GeForce RTX 50 series GPUS in UL's 3DMark Steel Nomad benchmark. ComputerBase community users with RTX 5070, 5070 Ti, and 5080 GPUs reported three to eight percent score uplifts after upgrading, prompting ComputerBase to verify these claims under controlled conditions. Their independent testing confirms that the RTX 5080 jumps from 8,094 to 8,550 points (a 5.6 percent boost), the RTX 5070 Ti climbs from 6,463 to 6,932 points (7.3 percent), and the RTX 5070 improves from 4,838 to 5,242 points (8.4 percent).”
 
chameleoneel said:
"Sleep" issues are usually driver and/or BIOS issues.
Click to expand...
Turn off the display(after XX min/hours) under Power options:
-I've been having problems with this lately, is it the Nvidia drivers or did Microsoft screw something up with the updates?

I've always used it, when I download something (e.g. a game) my monitor turns off after e.g. 5 minutes and I wake it up by simply moving the mouse, however it can be buggy, it turns off e.g. in a game and I can't wake it up, or it doesn't turn off completely but only partially (black screen but the lighting works)
Whose fault is that?
 
This newest driver brought back the "boot to black" issue. Windows loads but no display unless I unplug the cable and plug it back in to a different port. This is getting really, really old.
 
djoye said:
RTX 3090 here. With 572.83 I was getting random black screens in Atomic Heart, almost as if my GPU was overheating because the image would re-appear, but it didn't affect Ghost of Tsushima that I played for 40+ hours. I rolled back to 566.36(?) and Atomic Heart ran fine. Decided to update to 576.02 and Atomic Heart is doing well with no random black screens.

I think that having a DisplayPort and HDMI display connected was a condition that created a black screen, that's what I have. I won't be surprised if the issue returns.
Click to expand...
Just got the black screens back in Atomic Heart. Played a few times throughout the day without issue, then it appeared in a specific area. GPU wasn't even being hammered because I was in a confined space. Maybe I need to use DDU.
 
  • Like
Reactions: erek
like this
djoye said:
Just got the black screens back in Atomic Heart. Played a few times throughout the day without issue, then it appeared in a specific area. GPU wasn't even being hammered because I was in a confined space. Maybe I need to use DDU.
Click to expand...
Atomic Heart is the game that gave me my first crash too. It was the elevator shaft location but I haven't booted it up since then. I started reading about crashes I was on MSI 4080 then
 
I've been getting that issue where it's showing the card running at like sub-100 MHz speeds, but it doesn't seem to actually affect anything.
 
I had to roll back, every single time my PC boots up its a black screen until I unplug the monitor cable and plug it into a DIFFERENT port on the GPU. Can't be the same one, has to be different. Can't wait until my 9070XT rolls in later this week so I can try that.
 
djoye said:
Just got the black screens back in Atomic Heart. Played a few times throughout the day without issue, then it appeared in a specific area. GPU wasn't even being hammered because I was in a confined space. Maybe I need to use DDU.
Click to expand...
:(
 
Strange, I've been running the 566.45 hotfix driver (The hotfix after 566.36) and everything's been working as it should..

What's odd is Nvidia pulled the driver from their website, which I don't recall ever seeing them do before....

It's still on guru3d.com if anyone plays Ubisoft snowdrop engine games/Indiana Jones and the Great Circle
 
horrorshow said:
Strange, I've been running the 566.45 hotfix driver (The hotfix after 566.36) and everything's been working as it should..

What's odd is Nvidia pulled the driver from their website, which I don't recall ever seeing them do before....

It's still on guru3d.com if anyone plays Ubisoft snowdrop engine games/Indiana Jones and the Great Circle
Click to expand...
That's weird, the only driver I have heard of after 566.36 is the 5090 launch day driver. That's where the problems started.
 
Wow new driver is consistently showing less wattage and temps compared to older drivers. I am hovering around 350 watts in BO6 with a curve that is more aggressive than my last one -> 975 @ 2917 vs. 965 @ 2917. I played for hours and did not see it go above 400 watts. Previously it would hover between 450-475 watts. Temps wise also it stays under 65 C whereas, previously it would try to climb to 71 C.
 
Astral 5090 is gone and have been playing with a 5070 Ti (Aorus Master - went to my son's rig) and then now a 5080 (Vanguard SOC - Facebook score) - no issues with the latter two cards and this driver with Fortnite. With an OC on the 5080 I'm a very happy recession-proof gamer. :)
 
  • Like
Reactions: Niner
like this
sk3tch said:
Astral 5090 is gone and have been playing with a 5070 Ti (Aorus Master - went to my son's rig) and then now a 5080 (Vanguard SOC - Facebook score) - no issues with the latter two cards and this driver with Fortnite. With an OC on the 5080 I'm a very happy recession-proof gamer. :)
Click to expand...
Yeah a 5080 overclocked is a really damn good card. Very close to 4090 performance.

Tried out these drivers and they finally do seem good on my 4090. Been using the December 10th drivers, so its nice that these look solid.

Wife is saying things are good with these drivers on the 5080. This is finally a good sign!
 
KickAssCop said:
Wow new driver is consistently showing less wattage and temps compared to older drivers. I am hovering around 350 watts in BO6 with a curve that is more aggressive than my last one -> 975 @ 2917 vs. 965 @ 2917. I played for hours and did not see it go above 400 watts. Previously it would hover between 450-475 watts. Temps wise also it stays under 65 C whereas, previously it would try to climb to 71 C.
Click to expand...


“NVIDIA GeForce 576.02 driver said to be affected by GPU temperature sensor bug
Published: Apr 21st 2025, 12:19 GMT 37 Comments”

https://videocardz.com/newz/nvidia-...ffected-by-serious-gpu-temperature-sensor-bug
 
You must log in or register to reply here.
Back
Top