GTX 480 Video Card Drivers - Then vs Now

[LinusTech]

Now do this with AMD for a truly interesting comparison; although that probably won't happen because we all know that Linus has been getting sponsored more by Nvidia Linus is biased towards Nvidia. Really annoying actually. @nicklmg

 

NVIDIA owns 80 percent of the discrete graphics card market. This video addresses 80% of our audience and was less than half as much work as one that includes another graphics card just to drive home the same point.

 

And besides, I don't have any 5 year old AMD graphics cards lying around unfortunately, so we threw the GTX 480 on the bench and went with it.

 

Nothing to do with h8ing on AMD or whatever. We feature their stuff where it makes sense (lots of VRAM, surround gaming for example - 

)

[Syntaxvgm]

They don't even have to link any videos. All they have to post is "We have a new video on Vessel. Check it out." Even the title of the vid isn't neccessary. Or, you can just make an email subscription using a 3rd party site to spam everyone who signs up to check out the new vid. I dunno, these are my opinions, and I have no place to tell LMG what to do. 

 

I can only suggest them to have a smaller bias or favouring towards nVidia and feature AMD in their videos sometimes. Everyone knows that Amd supports their cards for longer than nVidia.

As a reference 970 owner I got it knowing I'd be shit out of luck in 2 generations lol

[Syntaxvgm]

NVIDIA owns 80 percent of the discrete graphics card market. This video addresses 80% of our audience and was less than half as much work as one that includes another graphics card just to drive home the same point.

 

And besides, I don't have any 5 year old AMD graphics cards lying around unfortunately, so we threw the GTX 480 on the bench and went with it.

 

Nothing to do with h8ing on AMD or whatever. We feature their stuff where it makes sense (lots of VRAM, surround gaming for example - 

)

I'd like to point out you could walk out with an 8350 as a belt buckle, a set of Armour made from AMD cards, and pair of nunchucks made from 295x2s and your viewers will say you have an NVIDIA bias.

[tedlas]

As a reference 970 owner I got it knowing I'd be shit out of luck in 2 generations lol

The 970 is not a bad choice, It's a great card. I would have just thought, of all topics to actually feature an Amd card in a comparison, this would be one of them. 

[Str_Mape]

I'd like to point out you could walk out with an 8350 as a belt buckle.

If that a reference to the belt buckle that came with a few of the 8150's? I feel like it is.

[Syntaxvgm]

If that a reference to the belt buckle that came with a few of the 8150's? I feel like it is.

lol I didn't know that was a thing.

That as a belt buckle and an 8350 as an eye patch then?

I mean it's PGA so that would hurt, but yah know

[Bob Jim]

NVIDIA owns 80 percent of the discrete graphics card market. This video addresses 80% of our audience and was less than half as much work as one that includes another graphics card just to drive home the same point.

 

And besides, I don't have any 5 year old AMD graphics cards lying around unfortunately, so we threw the GTX 480 on the bench and went with it.

 

Nothing to do with h8ing on AMD or whatever. We feature their stuff where it makes sense (lots of VRAM, surround gaming for example - https://www.youtube.com/watch?v=Bteq-6D1srs)

My point was more to see which company actually has better driver support in terms of performance, because the amount of people saying, oh well Nvidia/AMD has so much better driver support than Nvidia/AMD. Shame you don't have an old AMD card though.

[tedlas]

NVIDIA owns 80 percent of the discrete graphics card market. This video addresses 80% of our audience and was less than half as much work as one that includes another graphics card just to drive home the same point.

 

And besides, I don't have any 5 year old AMD graphics cards lying around unfortunately, so we threw the GTX 480 on the bench and went with it.

 

Nothing to do with h8ing on AMD or whatever. We feature their stuff where it makes sense (lots of VRAM, surround gaming for example)

I guess your videos aren't really helping to make the market shares more equal. 

 

However, Amd does support it's gpus better than nVidia does. Especially lately: http://www.hardwarecanucks.com/forum/hardware-canucks-reviews/70125-gtx-780-ti-vs-r9-290x-rematch.html

[Syntaxvgm]

I guess your videos aren't really helping to make the market shares more equal. 

 

However, Amd does support it's gpus better than nVidia does. Especially lately: http://www.hardwarecanucks.com/forum/hardware-canucks-reviews/70125-gtx-780-ti-vs-r9-290x-rematch.html

I wouldn't have agreed until the recent string of fucked drives I got from NVIDIA from the release of witcher 3 until recently. Least I wasn't a 980ti owner, I could actually use old drivers that worked. 

I've always thought AMD's drivers were worse, but better in principal. Better gpu support, didn't wait several years to support basic begged for features like hdmi full color range (and patch it out when people forced it in), ect

[Mew]

Would like to see this applied to a wider range of graphics cards to see how it affects the top of the line, mainstream, and entry level to see if they all show the same or similar gains. 

[Kirky2k15]

Anyone seen the voodoo 3 2000 (my 1st GPU) unboxing on pcper? Do this what that! lolz

 

Anyway, for the red team, my 290x went from 95xx in firestrike to 105xx with a driver update (no OC) so there is a good argument to do this for the red team as well IMO

[ybriK]

Really interesting but would love to see some love for the red team aswell. I also wonder how drivers have an impact on overclocking capabilities.

[NicekillBG]

LinusMasterRace

[gluconolactone]

-snip-

 

I heavily doubt that 80% of your audience own an Nvidia card. While it is probable that Nvidia card owners are more abundant over here, it wouldn't stretch to 80%. This is because people in this community are mostly PC Builders which make educated decisions with regards to purchasing their parts. Nvidia's 80% market share is aided by the fact that most prebuilt Dells and HPs have Nvidia GPUs preinstalled (GT 610 and GT 730 for example), and even within PC builders some buy parts without realizing their comparative performance and the general consensus in the public is that Nvidia GPUs are better across the board, so people seem to go with that. Over here I would guess the percentage is more in the 60-40 range though.

 

Not claiming any bias, just stressing the importance of making similar comparisons for AMD cards as well.

[Progressor]

Disappointing and even unprofessional video, especially for someone of your magnitude, guys.

 

1. The card you used is way too old, something more relevant like the gtx780 or 780ti would be better. I mean who cares how optimized your drivers would be after 5 years(seriously, 5 fucking years?), drivers are optimized withing weeks/month not years and when people say mature they usually mean that period of few weeks/month. Test of drivers up to 2 year after a card is released is more than enough, hence why you should used a newer card(and would allowed you to use AMD card if that's your excuse, *point 3 below). 

 

2. One month after a card is released drivers are already too old. As I said drivers are optimized withing weeks if not days after/during release so what you ended up testing(by starting with  1 month old drivers) is a bunch of already optimized drivers. You should have started with the first released drivers that adds support for the card you're testing(even if the driver come out before the card was released for sale).

 

3.  No AMD? I don't care how much % of people use AMD/Nvidia. AMD cards have enough people using them that they deserve testing just as much as Nvidia cards. If you really care about percentage and appealing to the "majority" of your userbase you should've made a strawpoll and ask you audience to vote(just because statistics or whenever you pulled that 80% from show that 80% of the general users use Nvidia, doesn't that it's the same percentage for your specific type of users that watch you). Also why cover only "80%" of users when you could easily covered all of them(use newer nvidia cards, or if you really don't want to then use old nvidia and newer amd cards, there is no excuse not to use amd considering your fanbase and the hardware resources your company has).

 

4. Put you tinfoil hats gentlemen. Now something else that would have been interesting to see is: does Nvidia/AMD cripple their drivers when a newer generation of cards is released, for example are 780/ti drivers crippled after 970/980/ti were released or the same with radeon 280/290/x and 390/Fury/whatever.

 

Now, with all that it would've been an interesting, objective video looking critically at all aspects of your topic(Then vs Now Video Card Drivers).

Sorry guys, but the weird way you did the tests and the video, it looks as if Nvidia threw  a bunch of cash at you and told you to make an ad for them(I know it sounds ridiculous and blaming but that's how it looks).

 

EDIT because replies:

My comment is targeted at Linus and the guys involved in this video and I don't really care that much about the opinion of my post by the other dudes here. I'm not big on posting in forums anyway. Although my initial post is negative, it's meant as a critique not hating so If anyone from LMG is reading it(I hope so, otherwise it's pointless), don't take it the wrong way. I've been a subscriber of LTT for awhile(and the other channels since they started), love the LMG team and what they do, except for that particular video which is the first LTT video I pressed dislike on and disagreed with to the point of going to the forum here and giving my 2 cents.

[shawno]

Id also like to see some amd stuff.

Getting ahold of an old amd card isnt going to be hard at all.

[antoainb]

My point was more to see which company actually has better driver support in terms of performance, because the amount of people saying, oh well Nvidia/AMD has so much better driver support than Nvidia/AMD. Shame you don't have an old AMD card though.

 

That would be a cool video, but it wasn't the purpose of the video, this was to see how much performance had changed not to see which manufacturer had better support.

 

I guess your videos aren't really helping to make the market shares more equal. 

 

However, Amd does support it's gpus better than nVidia does. Especially lately: http://www.hardwarecanucks.com/forum/hardware-canucks-reviews/70125-gtx-780-ti-vs-r9-290x-rematch.html

 

Why on Earth should their videos be balancing market shares? That's AMD's job to make better cards and market them, not Linus'. Besides, then he'd be biased towards AMD.

 

I heavily doubt that 80% of your audience own an Nvidia card. While it is probable that Nvidia card owners are more abundant over here, it wouldn't stretch to 80%. This is because people in this community are mostly PC Builders which make educated decisions with regards to purchasing their parts. Nvidia's 80% market share is aided by the fact that most prebuilt Dells and HPs have Nvidia GPUs preinstalled (GT 610 and GT 730 for example), and even within PC builders some buy parts without realizing their comparative performance and the general consensus in the public is that Nvidia GPUs are better across the board, so people seem to go with that. Over here I would guess the percentage is more in the 60-40 range though.

 

Not claiming any bias, just stressing the importance of making similar comparisons for AMD cards as well.

 

If you were running a BUSINESS, and had to make a video that was going to be a HUGE amount of work what would you do? Would you:

1. Test the product most relevant to your viewers and release the video since the results for other products are going to be very similar.
2. Run it on 2 products and find there is no real difference worth noting and have just wasted resources that could have been put towards another video, maybe of higher importance?

 

Disappointing and even unprofessional video, especially for someone of your magnitude, guys.

 

1. The card you used is way too old, something more relevant like the gtx780 or 780ti would be better. I mean who cares how optimized your drivers would be after 5 years(seriously, 5 fucking years?), drivers are optimized withing weeks/month not years and when people say mature they usually mean that period of few weeks/month. Test of drivers up to 2 year after a card is released is more than enough, hence why you should used a newer card(and would allowed you to use AMD card if that's your excuse, *point 3 below). 

 

2. One month after a card is released drivers are already too old. As I said drivers are optimized withing weeks if not days after/during release so what you ended up testing is a bunch of already optimized drivers. You should have started with the first released drivers that adds support for the card you're testing(even if the driver come out before the card was released for sale).

 

3.  No AMD? I don't care how much % of people use AMD/Nvidia. AMD cards have enough people using them that they deserve testing just as much as Nvidia cards. If you really care about percentage and appealing to the "majority" of your userbase you should've made a strawpoll and ask you audience to vote(just because statistics or whenever you pulled that 80% from show that 80% of the general users use Nvidia, doesn't that it's the same percentage for your specific type of users that watch you).

 

4. Put you tinfoil hats gentlemen. Now something else that would have been interesting to see is: does Nvidia/AMD cripple their drivers when a newer generation of cards is released, for example are 780/ti drivers crippled after 970/980/ti were released or the same with radeon 280/290/x and 390/Fury/whatever.

 

Now, with all that it would've been an interesting, objective video looking critically at all aspects of your topic(Then vs Now Video Card Drivers).

Sorry guys, but the weird way you did the tests and the video, it looks as if Nvidia thrower a bunch of cash at you and told you to make an ad for them(as ridiculous and blaming as it sounds).

 

1. The 7xx series is still very much heavily supported hence it wouldn't make much sense to test. It makes more sense to test an old card because it's not as heavily supported anymore and you can bet the results will be true for longer than with a card with a lot of support right now, which might change performance quite quickly.
2. That has nothing to do with the video other than the topic of drivers.
3. You might not care that more people use nvidia but they do and they should make their videos efficiently and in a way they are relevant to the maximum amount of people. Yes, covering AMD would have been inefficient since the results probably wouldn't have been interesting and it would have been a waste of time for the benchmarking team which is probably hard at work with Skylake already.
4. That has nothing to do with the video other than the topic of GPUs.

This video was objective since they did objective testing, it would have been subjective if they showed us the numbers and proceeded to make conclusions that didn't come from the numbers but from personal beliefs, much like how you want AMD coverage not because it makes sense but because.

This wasn't and ad for nvidia, they didn't f**k up drivers so they didn't say anything bad about them, but they didn't say anything that wasn't truth or say at any point that nvidia did something better than AMD.

[antoainb]

I've always been annoyed they don't make topics until it hits youtube, MAYBE this is their way around that. Notice it's not public on youtube yet? Unlisted. 

Could be wrong though 

 

They release some videos simultaneously on both platforms. They also make this post before the video is made public, but not at the same time it's on vessel, only for special videos.

[IMPERIUS]

Do the same for AMD gpus @LinusTech @nicklmg @Slick

[Centurius]

My point was more to see which company actually has better driver support in terms of performance, because the amount of people saying, oh well Nvidia/AMD has so much better driver support than Nvidia/AMD. Shame you don't have an old AMD card though.

 

That wasn't the intention behind the video though, it was to show that even years from purchase you can still have a better gaming experience than the day after purchase. That's pretty much brand-neutral and doesn't require both to be compared.

[TheArni]

The problem is, though, that drivers aren't "being optimized as the software engineers learn their hardware". First of all, I think the hardware engineering team and the software engineering team at NVIDIA are working together more often than not, so I'm pretty sure that at release the software engineers know well enough the hardware they're working with.

 

The real problem is that, more often than not, games come out as an unoptimized mess of code. What happens is that, NVIDIA specially (that's probably why AMD wasn't included in the video, because they don't have this working scheme) work very close to the game developers, and they try to fix their "errors" on the driver side, since they aren't actually allowed to look at the source code or modify it by any means.

 

And that's why we usually see "New NVIDIA Graphics Card update specially optimized for Some Game", simply because they are optimizing someone else's code so it doesn't run poorly on their hardware (and that's probably the reason why some par to par AMD graphics card have worse performance than their NVIDIA counterparts).

 

So, in conclusion, no, graphics card drivers don't magically squeeze performance out of a graphics cards over time. They just add game-dependant fixes that improve that particular game's performance.

[soaringchicken]

I don't even really care if an AMD video comes out or not to be honest. I highly doubt AMD vs Nvidia support is drastically different anyway. Besides the point, it's stupid to think people are complaining about a lack of something in free content. 

[gluconolactone]

I don't even really care if an AMD video comes out or not to be honest. I highly doubt AMD vs Nvidia support is drastically different anyway. Besides the point, it's stupid to think people are complaining about a lack of something in free content. 

 

Less of a complaint, more of a suggestion.

 

If you were running a BUSINESS, and had to make a video that was going to be a HUGE amount of work what would you do? Would you:

1. Test the product most relevant to your viewers and release the video since the results for other products are going to be very similar.
2. Run it on 2 products and find there is no real difference worth noting and have just wasted resources that could have been put towards another video, maybe of higher importance?

 

If I was running a business I would try to cater to as much of my audience as possible. I doubt it would have been very difficult to obtain say a 6790 and run the same tests.

I'm not trying to bash anyone, I just went into the video expecting to see an old AMD card and an old Nvidia card, as one typically would, since the YouTube title didn't particularly mention the 480.

[soaringchicken]

Less of a complaint, more of a suggestion.

 

 

If I was running a business I would try to cater to as much of my audience as possible. I doubt it would have been very difficult to obtain say a 6790 and run the same tests.

I'm not trying to bash anyone, I just went into the video expecting to see an old AMD card and an old Nvidia card, as one typically would, since the YouTube title didn't particularly mention the 480.

You obviously haven't looked at the youtube/vessel comments then, omg they're awful. 

[MahtXL]

This is also a great thing to apply to the current gen consoles (which are already being tapped out) "BUT BUT BUT, DEVS WILL LEARN THE SYSTEM AND THINGS WILL GET BETTER" the console players cry. No not this time around buds, its x86, theres not much to "unlock" or "learn" you get what you get. So i guess its been settled on all fronts. When you buy a piece of tech, console or gpu, take it at face value because it wont get much better.