Nvidia releases bug fix 364.51 drivers to address installation issues

[Scorpus]

Yesterday Nvidia launched a new set of drivers for their GeForce graphics cards, and it hasn't been the smoothest release so far. Many users have reported issues with the 364.47 build, particularly with the installer, causing Nvidia to quickly release a fix that you should install instead.

The main issue with the 364.47 drivers concerned its installation process, with some users experiencing a crash during the express installer that breaks their system. As the crash would then prevent affected users from booting into Windows, they'd have to either boot into safe mode and roll back to a previous driver version, or revert to a previous point using system restore.

Nvidia became aware of the issue shortly after users posted about their struggles on Reddit and other forums, and has released a new driver that includes a fix. Version 364.51 is still in beta at the moment, pending WHQL approval from Microsoft, but you should grab this version instead of 364.47 to ensure there's no issues during the installation process.

Aside from the fixes to the installer, the 364.51 driver is identical to 364.47 in its feature set.

On a less critical note, The Tech Report has discovered that the high-refresh-rate power bug that plagued earlier versions of Nvidia's GeForce driver has appeared once again, at least in version 364.47. Using an Asus PG279Q, a GTX 980 Ti, and a refresh rate of 135 Hz, the site managed to get the GPU idling at 925 MHz, which is well above the standard idle clock speed of 135 MHz.

The increased clock speed at higher refresh rate causes a pretty significant increase in idle power consumption. The test system used by The Tech Report rose from around 85W to around 140W due to the issue, which will have an effect on a user's power bill over time. Hopefully Nvidia can identify the issue and roll out an update in the near future.

Permalink to story.

https://www.techspot.com/news/64037-nvidia-releases-bug-fix-36451-drivers-address-installation.html

 

[Evernessince]

Good, now I don't want to hear either side complaining about driver issues. Both sides have had their fair share of driver mishaps.

 

[howzz1854]

Good, now I don't want to hear either side complaining about driver issues. Both sides have had their fair share of driver mishaps.

I must've brought it over from AMD. recently with the fiasco of Crimson driver, I finally got fed up and sold the crossfire and went with a 980 Ti, and now this.

but in all honesty, neither is GREAT or driver problem free. that's the inherent problem with drivers. it's a critical component within a system, it's easy and bound to go wrong from time to time. the only difference is, with AMD, they're slow to the rescue. with Nvidia, and their bigger budget, they have more engineers to bandage the damage faster.

 

[Evernessince]

I must've brought it over from AMD. recently with the fiasco of Crimson driver, I finally got fed up and sold the crossfire and went with a 980 Ti, and now this.

but in all honesty, neither is GREAT or driver problem free. that's the inherent problem with drivers. it's a critical component within a system, it's easy and bound to go wrong from time to time. the only difference is, with AMD, they're slow to the rescue. with Nvidia, and their bigger budget, they have more engineers to bandage the damage faster.

Nvidia is quicker but I think they need to be. They are releasing drivers for pretty much every game with the only benefit being to help with performance of that game. It'd be pretty bad if such a small update with such a large bug would remain for very long. Not hard to compare code when the difference is minute.

Personally I just ignore the Game Ready drivers and GForce experience all together. I didn't really notice a performance boost for any of the updates and for me it was more burdensome that helpful with the constant "Game Ready Drivers" pop-ups.

I wouldn't mind if AMD did do it as much as Nvidia so long as they let you opt-out. Personally though I'd rather they focus on making the driver as good as possible and not focus on singular games.

 

[cliffordcooley]

I'm still running version 361.75. It works for me, so no need in staying up to date.

 

[hahahanoobs]

nVIDIA recommends you don't install 364.47 or 364.51 and wait. 362.00 drivers are safe.

 

[tomkaten]

Installed 364.47 over the previous version (362.00) on my two Win 10 x64 machines, no issues.

Then again, both are single monitor setups, so maybe that's why. I'll skip this fix, cause they both work fine.

 

[Arturo]

I have been running into problems with Nvidia driver installation for around one year already. Now, to install every driver I need to perform clean install, which stops me from using profiles for games. Now this. Thank God, I hadn't installed it yet.

 

[mosu]

I must've brought it over from AMD. recently with the fiasco of Crimson driver, I finally got fed up and sold the crossfire and went with a 980 Ti, and now this.

but in all honesty, neither is GREAT or driver problem free. that's the inherent problem with drivers. it's a critical component within a system, it's easy and bound to go wrong from time to time. the only difference is, with AMD, they're slow to the rescue. with Nvidia, and their bigger budget, they have more engineers to bandage the damage faster.

What Crimsom driver fiasco? If 20% gain means fiasco, better read this about Nvidia latest and greatest drivers here: link (http://wccftech.com/nvidia-users-beware-latest-drivers-damage-pc/)

 

[Skidmarksdeluxe]

It did affect everybody, my installation went as smooth as it usually does.

 

[dividebyzero]

What Crimsom driver fiasco? If 20% gain means fiasco, better read this about Nvidia latest and greatest drivers here: link (http://wccftech.com/nvidia-users-beware-latest-drivers-damage-pc/)

Probably referring to the issues with Fury cards and downclocking and fan speed issues. Anyhow, I think you'll find that the "Nvidia driver kills graphics cards" was - unsurprisingly- guerrilla marketing by people allied to another brand. The same site you just quoted said as much (fairly quickly exposed once the posting history of the people supposedly having dead cards was analysed):
NVIDIA’s Latest GeForce Drivers Are Not Burning or Frying Graphics Cards, Issues Mostly Related To Glitchy Displays and Crashes

As has already been noted here and elsewhere, neither driver team is covering itself in glory at the moment. Between a new Windows Display Driver Model, new API's and their attendant features, and a batch of new (half baked) games there are too many moving parts for the driver teams to fully account for.

It did affect everybody, my installation went as smooth as it usually does.

I'm assuming you meant "didn't" rather than "did". No problems here (note: single monitor user) - but I driver sweep and fresh install - along with just installing the bare driver (graphics+HD audio+physx). Then it is just a matter or reapplying my OC settings and custom settings via Nvidia Inspector. No biggie.

 

[Puiu]

Proving that WHQL means nothing when it come to assuring quality.

 

[robb213]

Proving that WHQL means nothing when it come to assuring quality.

Considering SLI in The Division under Win10 causes a VRAM leak, a la the DSR VRAM leak that was a problem a few months ago.

Even though I can imagine how complicated these drivers are, QA seems to be falling.

 

[amstech]

Good, now I don't want to hear either side complaining about driver issues. Both sides have had their fair share of driver mishaps.

This is absolutely true, even though AMD has had about 4-5 times as many issues every year for the past 10 years, Nvidia is not immortal by any means.
And hail Scorpus, the great AMD scholastic and professional AMD fan.

 

[Evernessince]

This is absolutely true, even though AMD has had about 4-5 times as many issues every year for the past 10 years, Nvidia is not immortal by any means.
And hail Scorpus, the great AMD scholastic and professional AMD fan.

I'd like to see either of your statements proven. Otherwise, you are the only to to mutter "fan" anything in this thread.

 

[howzz1854]

What Crimsom driver fiasco? If 20% gain means fiasco, better read this about Nvidia latest and greatest drivers here: link (http://wccftech.com/nvidia-users-beware-latest-drivers-damage-pc/)

The infamous Fan issue where the fan sets itself to 0% and overheating graphics cards. the new Qt back end caused the state of graphic to be in performance mode 100% of the time. per game profile don't always work. infamous flickering issue related to refresh rate either in game or on desktop. all were NOT present during catalyst, but brought on by Crimson. the progress in rectifying those issues are very slow. luckily the fan problem was fixed much quicker. clearly they rushed the release of crimson. I would've wished they simultaneously release crimson and catalyst and kept catalyst going for a little while longer while people test out the new crimson.

 

[howzz1854]

I thought I'd report back that after yesterday's posting. I went home and installed 364.51. mind you that 364.47 installed fine for me the night before. but a friend suggested that I installed the new 364.51 to be sure rid of 364.47. and what do you know, I ran into the exact black screen issues.

here is my write up on the Nvidia forum.
https://forums.geforce.com/default/...thread-released-3-7-16-/post/4829930/#4829930

but long story short, the installer/uninstaller hosed the Displayport connection. there's nothing you can do to bring it up back. only way I was able to get the rig back online is through a legacy connection DVI and an old monitor that was 1920x1080 wit a dvi port. had to install the driver with the legacy port to get the system to recognize DP again. not only that once it recognized it, you swap the DVI back to DP and boot it back up, it fails to see the video adapter, but only shows it as Generic VGA adapter. I had to install the driver on top of it again and boot it back up for it recognize my 980ti. it's a freaking mess. clearly Nvidia has no understanding what's really causing the 364.47 and just blindly rushed out a new driver which was just the same or as bad.

I don't plan on going back to the red team anytime soon, because their Crimson driver is also having all sorts of issues, fan problem was only one of them, and hopefully was the last serious issue they'll have. what we really need is a third graphic manufacture in the mix.

 

[Evernessince]

I thought I'd report back that after yesterday's posting. I went home and installed 364.51. mind you that 364.47 installed fine for me the night before. but a friend suggested that I installed the new 364.51 to be sure rid of 364.47. and what do you know, I ran into the exact black screen issues.

here is my write up on the Nvidia forum.
https://forums.geforce.com/default/...thread-released-3-7-16-/post/4829930/#4829930

but long story short, the installer/uninstaller hosed the Displayport connection. there's nothing you can do to bring it up back. only way I was able to get the rig back online is through a legacy connection DVI and an old monitor that was 1920x1080 wit a dvi port. had to install the driver with the legacy port to get the system to recognize DP again. not only that once it recognized it, you swap the DVI back to DP and boot it back up, it fails to see the video adapter, but only shows it as Generic VGA adapter. I had to install the driver on top of it again and boot it back up for it recognize my 980ti. it's a freaking mess. clearly Nvidia has no understanding what's really causing the 364.47 and just blindly rushed out a new driver which was just the same or as bad.

I don't plan on going back to the red team anytime soon, because their Crimson driver is also having all sorts of issues, fan problem was only one of them, and hopefully was the last serious issue they'll have. what we really need is a third graphic manufacture in the mix.

I think that Nvidia knows what is causing the issues but for people who already installed the botched drivers, I believe they made things worse. It's like cracking a rock and then filling it in with putty. For users who've already had their rock cracked a simple patch isn't going to be enough.

 

[howzz1854]

Considering SLI in The Division under Win10 causes a VRAM leak, a la the DSR VRAM leak that was a problem a few months ago.

Even though I can imagine how complicated these drivers are, QA seems to be falling.

it's interesting that you brought up QA. I don't think the both sides test their drivers enough in Displayport format. there're way too many problems user reported with DP and HDMI. if they want DP to be the standard for today/tomorrow, they really need to test DP as the primary platform from now on.

 

[howzz1854]

nVIDIA recommends you don't install 364.47 or 364.51 and wait. 362.00 drivers are safe.

where did they mention that you skip 364.51 as well. because that will be a good idea consider what I had to go through yesterday.

 

[robb213]

it's interesting that you brought up QA. I don't think the both sides test their drivers enough in Displayport format. there're way too many problems user reported with DP and HDMI. if they want DP to be the standard for today/tomorrow, they really need to test DP as the primary platform from now on.

Yeah, I'm still surprised by DP alone. I convinced a friend to go with DP with his 980 Ti's, and it's only been a problem for him. Just told him to go back to HDMI since he was originally using that. Among all the problems I see people post about, HDMI and DP seem to be the most common each release for Nvidia (not sure about AMD since I don't prowl on there).

Me on the other hand, I still use DVI-DL. So I have no problems with I/O ever...thankfully