Wednesday, March 9th 2016

NVIDIA Pulls GeForce 364.47 WHQL Driver Over Critical Bugs, Outs 364.51 Beta

NVIDIA today pulled its latest GeForce 364.47 WHQL drivers for shipping with a critical bug in its installers that causes Windows to crash. Users swarmed to the company's official GeForce Forums to report the bug to NVIDIA. Among the problems included a crash each time the machine wakes up from Sleep mode. A company rep responded that NVIDIA was able to internally reproduce the issue, and was working on a solution.

NVIDIA has meanwhile pulled 364.47 WHQL from its website, and is recommending users to either clean-install an older version of the driver, such as the 362.00 WHQL, or upgrade to the newer 364.51 Beta Drivers. Version 364.51 Beta can be downloaded from the links below. The rest of its feature-set appears to be identical to the 364.47 WHQL.

DOWNLOAD: NVIDIA GeForce 364.51 Beta for Windows 10 64-bit | Windows 10 32-bit | Windows 8/7/Vista 64-bit | Windows 8/7/Vista 32-bit
Add your own comment

66 Comments on NVIDIA Pulls GeForce 364.47 WHQL Driver Over Critical Bugs, Outs 364.51 Beta

#26
Fx
EarthDogIts not a tale... don't let your experience (a pico sized drop in an ocean of experiences) dictate reality. BOTH have had problems at some points, but its pretty well known that AMD has more of their share.

He also didn't say "riddled" with problems.... "their share" is what was said. :)


Anyhow, gents, this is about NVIDIA drivers... don't turn this into a pissing match like would happen at OCN. ;)
10 years of multiple cards, multiple updates and multiple applications/games isn't pico-sized. There is no way you can define that as an isolated case or a lucky one-off.
Posted on Reply
#27
EarthDog
Yes sir, it is. You are one of TENS(hundreds?) OF THOUSANDS OF PEOPLE that have used AMD drivers over the past 10+ years. Pico brother... PICO sample size.

Now, move on or PM me... Im not getting in trouble answering this nonsense again. ;)
Posted on Reply
#29
Casecutter
I'd say the presumption of money/resources Nvidia and the owner base have used to form such a reputation is not holding up. This and several indiscretions within the last year reveal sizeable cracks in such a hypothesis.
Posted on Reply
#30
the54thvoid
Super Intoxicated Moderator
CasecutterI'd say the presumption of money/resources Nvidia and the owner base have used to form such a reputation is not holding up. This and several indiscretions within the last year reveal sizeable cracks in such a hypothesis.
There is no such reputation to hold up. Several years ago cards were fried when an Nvidia driver release stopped fans spinning. The only people that 'scream' about Nvidia's flawless drivers are:

A) Nvidia Fanboys and
B) AMD fanboys seeking to point out there are no good Nvidia drivers.


AMD drivers are well known to have had issues with select users. So much so, even as an Nvidia user (for now) I am aware many AMD folks that post about drivers always do clean installs because over installs usually have poor results.

But to suppose there are no Nvidia problems is just naive. Everyone should really stop being so weird about this. It's an abhorrent clamouring of frenzied finger waving like a bitter religious zealot pointing out another religion's inadequacies and it's frankly ridiculous. It's a driver issue - deal with it. Nvidia has them as does AMD.

FWIW - as an AMD user with 5850's and 7970's - I had zero problems with drivers, or none that were serious. Also never had a serious Nvidia problem either.

Finally, is this just aW10 issue because it could well be a M$ issue trying to scupper any Vulkan foothold to threaten it's beloved DX12. Or is that too conspiracy?
Posted on Reply
#31
Static~Charge
btarunrNVIDIA today pulled its latest GeForce 364.47 WHQL drivers for shipping with a critical bug in its installers that causes Windows to crash.
WHQL? More like WTF.... :(
Posted on Reply
#32
Xzibit
the54thvoidFinally, is this just aW10 issue because it could well be a M$ issue trying to scupper any Vulkan foothold to threaten it's beloved DX12. Or is that too conspiracy?
Reports are wider then just W10 and Vulkan. By the updates it seems they are just guessing at whats wrong with something that never should have made it out.
Nvidia ForumsIssues and Updates

7:30AM 3/8/16
After finding a critical installation issue, the team has replaced yesterday’s driver (364.47) with today’s new driver (364.51). This driver has been submitted to Microsoft for WHQL-approval and we will update the driver package online as soon as we have the certified package.



4:00PM PST
We have reproduced the issue and our engineers continue to debug it. Initial investigation suggests the issue is related to doing an "Express" installation. We suggest those willing to try this driver use a Custom -> Clean Install installation.


9:30AM PST
We have received reports of some users having issues installing today’s Game Ready Driver. Initial investigation suggests the issue is isolated to multiple-monitor configurations.
Should you experience issues, you can either disconnect one monitor during the driver installation (and reconnect once completed) or you can uninstall the driver through Windows Safe-Mode and revert to a previous version. Our driver team is actively investigating the issue and we will have more information shortly.
Conspiracies. Its Nvidia try'n get rid of old Kepler and Maxwell inventory through a driver. Just happens the inventory they are counting is in users PCs. Like the slow driver deteriation of Kepler when Maxwell was launched. Might be a sign of Pascal being announced soon. Que the X-Files theme music.
Posted on Reply
#33
Casecutter
the54thvoidEveryone should really stop being so weird about this.
Oh I acquiesce... that’s why I used the word "presumption".

All the... “one manufacturer bests the other" really became old news back at the RV870 (2009) era. That said there’s always mumblings to maintain just such a perception.
Posted on Reply
#34
R-T-B
MusselsI wish this happened back when people even made news articles over AMD not having WHQL for that 9 month period, because as is now abundantly clear - WHQL means squat.
I think AMD needs to recognize this too and stop labeling WHQLless drivers "beta". That has always been my chief complaint with their policy.
Posted on Reply
#35
qubit
Overclocked quantum bit
This explains why I wasn't able to download it from the geforce.com just a day after it went live. I'm glad I couldn't.
Posted on Reply
#36
GhostRyder
qubitThis explains why I wasn't able to download it from the geforce.com just a day after it went live. I'm glad I couldn't.
I got it, though I didn't have any problems and it fixed my machines one issue. Though I didn't try putting my laptop to sleep so I dunno.
R-T-BI think AMD needs to recognize this too and stop labeling WHQLless drivers "beta". That has always been my chief complaint with their policy.
Yea, that might be a good idea actually.
the54thvoidThere is no such reputation to hold up. Several years ago cards were fried when an Nvidia driver release stopped fans spinning. The only people that 'scream' about Nvidia's flawless drivers are:

A) Nvidia Fanboys and
B) AMD fanboys seeking to point out there are no good Nvidia drivers.


AMD drivers are well known to have had issues with select users. So much so, even as an Nvidia user (for now) I am aware many AMD folks that post about drivers always do clean installs because over installs usually have poor results.

But to suppose there are no Nvidia problems is just naive. Everyone should really stop being so weird about this. It's an abhorrent clamouring of frenzied finger waving like a bitter religious zealot pointing out another religion's inadequacies and it's frankly ridiculous. It's a driver issue - deal with it. Nvidia has them as does AMD.

FWIW - as an AMD user with 5850's and 7970's - I had zero problems with drivers, or none that were serious. Also never had a serious Nvidia problem either.

Finally, is this just aW10 issue because it could well be a M$ issue trying to scupper any Vulkan foothold to threaten it's beloved DX12. Or is that too conspiracy?
Yep, expecting a flawless driver 100% of the time is crazy. Not every situation out there is going to be accounted for and every once in awhile one gets missed. The only times to fret are when they physically kill your card (But that's a rarity).
Posted on Reply
#37
EarthDog
If you do not have multiple monitors, and/or installed it with the "clean" option, you shouldn't have had this issue.

People are making a lot of this... 'dat spilled milk... 'dat crying. :(
Posted on Reply
#38
RCoon
nem
Your last four posts have been irrelevant nonsense. I am tired of your repetitive asinine posts. Next time you earn moderator brownie points. Consider it your first and final warning.
Posted on Reply
#39
qubit
Overclocked quantum bit
GhostRyderI got it, though I didn't have any problems and it fixed my machines one issue. Though I didn't try putting my laptop to sleep so I dunno.
I use sleep mode a lot so I'd have probably been nailed by this problem, lol.
Posted on Reply
#40
EarthDog
Sorry... did this affect sleep?
Posted on Reply
#41
Solidstate89
EarthDogOther users did. It was there...

.. or maybe w10 forced it...
I'm on Windows 10. Never got it automatically.
Posted on Reply
#42
NightOfChrist
rtwjunkieThis is why I stick with a good driver for a good bit, and only upgrade to a new one (but not the latest), after I have seen it working well for awhile with other people.
True. I am still using Windows 8.1 Pro and the Always Ready driver never failed me, not even once. Come to think of it, it is almost a year old, but surprisingly - even to me - it works like new. I have tested it with many games, from Witcher 3 to Fallout 4, MGS: The Phantom Pain to Dragon's Dogma, Rise of the Tomb Raider to XCOM 2, and I never had a single problem with it. Photoshop, Lightroom and rendering with IRay work without any kind of problem too. The only downside is that driver does not have SLI profile for newly released games so I have to use just one GTX 980 instead of two. But it is an excellent driver. No problem for almost a year.

If only NVIDIA could release a driver like that again, particularly for Windows 10 users who I think have the bigger share of problems because of drivers...
Posted on Reply
#43
95Viper
btarunrNVIDIA has meanwhile pulled 364.47 WHQL from its website, and is recommending users to either clean-install an older version of the driver, such as the 362.00 WHQL, or upgrade to the newer 364.51 Beta Drivers.
I really don't get it why people whine and moan, about a driver being screwed up... they fixed it.
I am sure Nvidia and Microsoft did not try to do this to upset their customers. It happens, could have been a dozen+ reasons.

And, the feuding thing between camps is hilarious... it is like a bunch of 6 year olds on the playground.
I, myself, use both brands of cards... both are good; I have had drivers not install properly and had to fix them or roll back.
Ain't no big deal.

Now, if the driver actually damaged something, that is different, and they should remedy the situation... not saying they would, but, should.

Quit the bickering and post useful info... like, your experience with these drivers, if you had one.
The topic is not your experiences years ago or just to post to hear yourself gripe about what happened years ago.

Back on topic:

I did not get the pleasure of trying the 364.47 WHQL driver... I was late to the game.
However, I have installed the 364.51 Beta on my Windows 10 system, with the GTX960, upstairs and everything went smoothly, and have had no problems, so far.
I have to say thanks to NVidia for being quick to fix the situation.
Posted on Reply
#44
Prima.Vera
AssimilatorThis is a massive and unacceptable lack of quality control on nVIDIA's part. Truly disappointing considering how good and frequent their releases have been for the past year. Now is not the time to drop the ball.
Actually NO. In this case most of the fault is on Microsoft side because they issued the certification without proper testing. And nVidia is paying them a lot to do that.
Posted on Reply
#45
mcraygsx
btarunrLesson: WHQL doesn't mean jack...
exactly what I was thinking about.
Posted on Reply
#46
cushman
even the new beta has bugs and crashes just the same! smdh!!
Posted on Reply
#47
john_
btarunrLesson: WHQL doesn't mean jack...
I can point to an editorial that says the opposite. It goes even further
EarthDogWHQL does not mean that a driver will be trouble free guaranteed... is this a tech forum or a gardening forum? We should (all) know better!!!

en.wikipedia.org/wiki/WHQL_Testing
When a year ago people where saying that AMD's beta drivers are as stable or better than WHQL drivers, Nvidia fans where having a totally different opinion. Now that things turned around I am reading about tinfoil hats, sh!t happens and gardening forums. Well, have you posted that gardening comment in last years editorial?

I have to admit that I didn't knew and still don't know what Microsoft tests before giving that WHQL title to a driver. But you also have to understand that either this is a gardening forum and last year's editorial was written by an experienced gardener, or people who where saying ONE YEAR AGO that WHQL means nothing, and the only way to see a drivers quality and stability is by installing and testing it, where right.

One year ago everyone who was NOT a gardener was praising WHQL stamp. Today if you give as much credit to the WHQL stamp, as much as people where giving it last year, you are probably a gardener. Mine is mine and yours are also mine.
Posted on Reply
#48
EarthDog
@john_ , I can see where you take exception to last year's editorial compared to this thread and that post (for the record, I did not recall that thread and did not post in it - I checked). I can't speak for Btarunr, he would have to chime in and clarify. I can tell you, however, that I do not agree with that sentiment. I believe it to be shortsighted at best, and uninformed at worst. It is true that WQHL SHOULD be more stable as there SHOULD be more testing going into it. Is it perfect? Far from it as we have seen from BOTH camps. But its better than nothing (jack). At least it guarantees some form of additional testing. People need to understand that a poor driver release hurts whoever releases it, so it behooves them to go through this process, flawed or not.

As I said earlier, and appears you need reminded of, the tinfoil hat comment I made had everything to do with two specific posts about bribery and money on what WAS a $250 process that doesn't even exist anymore. If you read into any more deep than that, you putting words in my mouth. Same thing with the WHQL comment... you are pointing at and quoting the wrong person bub, there is no fanboy here. I would have posted that same sentiment in his editorial had I run across it last year. Take my statements at face value and leave me out of this pissing match between you and those that are actually fanboys. ;)
Posted on Reply
#49
john_
Just pointed at, how people change there opinions based on which company did what. It is really something that I see it very often. For things that people will never give any excuses for AMD, will throw a dozen excuses when it will concern Nvidia or Intel. It is nothing strange to be fair. Once someone tried to prove to me that this is not the case. With his post proved all my points.

As for that old thread, the editorial, you did liked one post there(just do a quick search with your nickname in the first page, haven't checked the other pages, let's not be paranoid), so you did see that one, but you didn't participate. Never mind. No one can be at every post.
Posted on Reply
#50
Mussels
Freshwater Moderator
He pointed out the same things i did, about how the attitudes have shifted now that Nvidia is the 'guilty' party, much like i did.

He just was not very polite about it. Bad John, be nicer.
Posted on Reply
Add your own comment
Nov 24th, 2024 09:50 EST change timezone

New Forum Posts

Popular Reviews

Controversial News Posts