Nvidia Fixes Their Driver Issues (it's not just AMD who has driver issues)

[Pickles von Brine]

Why is this being compared? You write code of ANY kind or you design ANYTHING and it be exploited one way or another. Some things just require people to be more clever. I am sorry, but other than the severity I hardly see why this is even being clickbaited the way it is. 

[LAwLz]

2 hours ago, Blademaster91 said:

Nah its needed because a lot of people on this forum like to think Nvidia has perfect drivers, which they don't, other issues with Nvidia drivers, with version 460.89 for example is screen flashing on 1080ti cards, youtube video stuttering when scrolling in a browser, or lagging in Steam VR. AMD would get tons of crap for things like screen flashing, but with Nvidia that seems to be mostly ignored.

I think it's the exact opposite. Nvidia gets shit for anything and everything they do, while AMD gets a pass on a lot of crap because "they are the little guy". At least on this forum. 

If you don't believe me, look at any thread about supply issues for the 30 series before AMD launched their graphics cards and then look at the threads about the RX 6000 series supply issues.

When the 30 series had issues - LOL Nvidia are so bad. It's a paper launch! It's a conspiracy theory. Nvidia are doing it on purpose!

When the RX series had even worse supply issues - Well it's Corona and it's understandable. Oh we should blame scalpers! Yeah, it's not AMD's fault. It's scalpers!

 

 

2 hours ago, Blademaster91 said:

Why not?

Because there are around 500 public vulnerabilities discovered and patched every month. Imagine if we had 20 threads EVERY DAY saying "product X has vulnerability Y".

For example The december update for AMD's drivers had 11 known issues in that driver alone, and it fixed 5 issues for previous drivers. Do we really need a thread saying "Latest AMD Driver Update Causes These Issues!" every month just because "AMD fanboys needs to get shown how bad AMD are"? I think that would just contribute to this terrible tribal mentality people has where they feel the need to "prove that company X is bad and company Y is better!".

[igormp]

10 hours ago, linuxChips2600 said:

As hinted by my title and intro, this article caught my eye especially as this was the first time I stumbled upon current news about Nvidia fixing serious driver issues

Nvidia had their drivers broken on kernel 5.9 for almost a month due to their shitty gpl-condom, users had to downgrade and not update until it was fixed, so driver issues aren't really that rare 

 

As for the bugs mentioned, most of those are not really important for mainstream consumers, but it's a serious issue for entreprise ones.

 

Anyhow, my package manager already has the latest driver since yesterday, so no worries for me.

[linuxChips2600]

6 hours ago, ragnarok0273 said:

Crap.

My GeForce 4 is not safe.

Sorry to hear that; I felt the same way when I was still carrying around a Samsung S4 mini not too long ago and watched Samsung patch more and more security flaws for their newer phones while dropping support for more and more of their older phones.  My personal experience with Samsung is why I switched to Pixel phones, given how Pixel phones have general received much better software support than their Android alternatives (with Google going even as far as allowing Pixel users to unlock their bootloader without fears of voiding their warranty).  However, with Google's recent (and imo regressive) trends of shortening their "guaranteed updates" periods for their phones and even just not bothering anymore to put the latest SOCs and hardware into their newest mobile products, my next phone after my Pixel 3 (although I'm not upgrading phones anytime soon) may not be a Pixel anymore.

Anyway, let me get back to the topic at hand and offer you some potential solutions/workarounds, if you don't mind:

1. Switch to Linux and use nouveau instead of Nvidia's proprietary legacy driver, knowing that you won't get even near the full 100% performance of your GeForce 4 (especially if you don't need the performance anyway).

2. Make sure to have the latest security/anti-malware software (preferably from a reputable vendor such as Norton or ESET) with the latest malware-definitions (or whatever the security companies call it these days) installed on the same machine you have your GeForce 4 on, minimizing any potential risk of others exploiting the same vulnerabilities that Nvidia just patched when you connect to the Web (especially given that these vulnerabilities are public knowledge now)

3. Upgrade your graphics card to one which Nvidia will continue to guarantee security updates for the foreseeable future.  By no means does it have to be a shiny, brand-new card, but preferably it's not something a bit too old like a GTX 450, so that you won't run into the issue of Nvidia dropping support for the the card you just upgraded to anytime soon.

Hope this helps, and I sincerely wish your computing experience to be pleasant and secure moving forward, even if it may not have been up until now.

 

[WhitetailAni]

Just now, linuxChips2600 said:

Sorry to hear that; I felt the same way when I was still carrying around a Samsung S4 mini not too long ago and watched Samsung patch more and more security flaws for their newer phones while dropping support for more and more of their older phones.  My personal experience with Samsung is why I switched to Pixel phones, given how Pixel phones have general received much better software support than their Android alternatives (with Google going even as far as allowing Pixel users to unlock their bootloader without fears of voiding their warranty).  However, with Google's recent (and imo regressive) trends of shortening their "guaranteed updates" periods for their phones and even just not bothering anymore to put the latest SOCs and hardware into their newest mobile products, my next phone after my Pixel 3 (although I'm not upgrading phones anytime soon) may not be a Pixel anymore.

A GeForce 4 is an AGP card.

It's in a Pentium III system that's not connected to the Internet via anything.
Also, I have an S4 Mini as a camera, Among Us, and flashlight (which I have to take a video to do) device.

[linuxChips2600]

2 minutes ago, ragnarok0273 said:

A GeForce 4 is an AGP card.

It's in a Pentium III system that's not connected to the Internet via anything.
Also, I have an S4 Mini as a camera, Among Us, and flashlight (which I have to take a video to do) device.

Ah I see; unfortunately given my age, to me AGP is a distant relic of the past (a.k.a. the standard was introduced before I was even born), so I didn't consider such old hardware when writing up my previous post.  I also see that you have made the wise choice of disconnecting your Pentium system from the internet.  I wish too that I could say the same as you of the longevity of some of the hardware I've owned, but unfortunately I may or may not have abused some of the electronics I've owned up until now...

Anyway, I'm still leaving up my initial reply to you in hopes that it may serve to be helpful to at least someone, if that's alright.

[Mark Kaine]

Nvidia drivers have their fair share of bugs, the difference is that they're usually not 'game breaking' just a bit annoying... and even those security bugs aren't anything new, they release fixes for those regularly, so yeah it probably wasn't necessary to mention AMD, especially because it's not the same type of 'issue'. 

 

 

11 hours ago, leadeater said:

Yep, much like the RTX 30 series issues that were actually drivers not the whole "BLAME THE CAPS!" like everyone was screaming 

 

Wait, it wasn't THE CAPS after all?? 

 

I knew they fixed it with drivers, but I did also believe it was to fix a "cap issue" in the first place.

 

(I was *very* tempted to not update drivers at all when I got my 3070 tbh, but I guess I'm glad I did, it's stable as heck too...!) 

 

Kind of strange everyone thought it was "the caps"... so generally speaking, it was the (transient*) power spikes? 

 

*tbh I looked up transient power spikes... but I didn't really understand it, it was all very technical, and from the description didn't seem to have to do anything with GPUs either lol... 

 

[WhitetailAni]

6 minutes ago, linuxChips2600 said:

Ah I see; unfortunately given my age, to me AGP is a distant relic of the past (a.k.a. the standard was introduced before I was even born), so I didn't consider such old hardware when writing up my previous post.  I also see that you have made the wise choice of disconnecting your Pentium system from the internet.  I wish too that I could say the same as you of the longevity of some of the hardware I've owned, but unfortunately I may or may not have abused some of the electronics I've owned up until now...

Anyway, I'm still leaving up my initial reply to you in hopes that it may serve to be helpful to at least someone, if that's alright.

I didn't disconnect it.

I have no way of connecting it.

Also, AGP is older than me too.

 

[linuxChips2600]

24 minutes ago, Mark Kaine said:

*tbh I looked up transient power spikes... but I didn't really understand it, it was all very technical, and from the description didn't seem to have to do anything with GPUs either lol... 

TBF "transient power spikes" is not a concern to the majority people out there, unless if you are an Electrical Engineer/Circuits Designer/etc.  But actually they *do* have a lot to do with GPUs, in the sense that GPUs (just like anything that requires electricity to run) incorporate various electrical circuits, and for those who go into the Electrical/Computer Hardware Engineering field for their college degree, one of the first engineering problems they are introduced to is how to design a circuit with transient voltage/power/amperage spikes in mind.  Then again, if we were all Electrical Engineers we probably won't need companies to produce consumer electronics for the rest of us, do we?

[DrMacintosh]

15 hours ago, dizmo said:

well, I assume AMD does

Yep, they release an update every month. 

[leadeater]

2 hours ago, linuxChips2600 said:

Ah I see; unfortunately given my age, to me AGP is a distant relic of the past (a.k.a. the standard was introduced before I was even born)

 

2 hours ago, ragnarok0273 said:

Also, AGP is older than me too.

Can we give a shout out to all those that have used multiple AGP GPUs? 

 

[StDragon]

24 minutes ago, leadeater said:

 

Can we give a shout out to all those that have used multiple AGP GPUs? 

My favorite AGP card at the time was a TNT2 Ultra used to play Quake 3 like a bat out of hell!

But that moment of "awe" in PC gaming bliss was the original Voodoo PCI card (Diamond Monster 3D).

 

 

[leadeater]

2 minutes ago, StDragon said:

But that moment of "awe" in PC gaming bliss was the original Voodoo PCI card (Diamond Monster 3D).

Well my PC usage predates PCI but I didn't really want to say that lol.

[williamcll]

But will Nvidia ever fix how their older GPUs tends to perform worse over subsequent driver updates?

[Moonzy]

Just now, williamcll said:

But will Nvidia ever fix how their older GPUs tends to perform worse over subsequent driver updates?

does it? i thought someone tried it and find that it was false rumor

 

though it was back in 2018

 

[Nowak]

The title made me think Nvidia finally fixed some longstanding bugs with their drivers, like the card idling at 1GHz in a mixed refresh rate setup. Disappointed to learn it's not that, but instead some security fixes.

 

Poke me when Nvidia does that.

[leadeater]

1 minute ago, Moonzy said:

does it? i thought someone tried it and find that it was false rumor

I think a few places looked at this and found it wasn't the case. What happens is the potential performance of Nvidia GPUs is attained from release, Nvidia includes game optimizations for supported GPUs, Nvidia stops supporting GPUs so GPUs no longer supported start to lose relative performance to their generational replacement so it looks like newer Nvidia drivers are regressing performance which is not the case. If you look at games that still have the optimizations for the older GPUs (which don't get removed for a very long time) the performance is the same.

 

Where as with AMD their GPUs almost never attain maximum potential performance at release and because of how much they rebrand and rerelease cards with old architectures older GPUs get game optimizations a bit longer than Nvidia GPUs, and as games get more complex and demanding AMD GPUs handle that slightly better. But we are talking very long time frames and by that point you should have replace the GPU anyway, so unless you are keeping your GPU for a VERY long time this doesn't matter.

 

Looking at performance over generations using current generation games of the time doesn't and cannot prove their is performance regression of older GPUs with each newer driver.

[WhitetailAni]

54 minutes ago, leadeater said:

Can we give a shout out to all those that have used multiple AGP GPUs? 

Thanks!

[Fnige]

4 hours ago, leadeater said:

 

Can we give a shout out to all those that have used multiple AGP GPUs? 

 

Was AGP more annoying than PCIe?

[leadeater]

9 minutes ago, SlimyPython said:

Was AGP more annoying than PCIe?

Nah just another slot, although you only got one of them so there were no SLI/Crossfire AGP cards. Also cards back then were away smaller so much easier to put in and take out.