Nvidia ready to counter AMD`s Mantle

[bystander36]

i don't know that he said that. is there broad consensus that he meant that? i dont know why he wouldve stayed with opengl if dx was easier. he doesnt seem to me like someone who would like uniform standards and the restrictive pipeline of dx11 but i could be wrong on any or all of that... im going to try to be open.

and no dx sucks because it has helped make nv and amd the only players since 2002 via its min specs that really picked up with the R300... the geforce fx didnt really suck that much more than the R300 because the latter had partial precision for the pixel shaders and it was a step back from the 8500 in terms of filtering and the 8500 even had 32 bit z-buffer support.
anyway, please don't be so satisfied with everything

also, if they're linking it to the Xbox One, then it won't ever not suck.

A simple Google search found this: http://www.bit-tech.net/news/gaming/2011/03/11/carmack-directx-better-opengl/1

 

[Paul98]

Yeah like I posted earlier in this thread, they slowplay each other in sake of profit.

Nvidia could have unleashed the GK110 from the getgo and be done with it. But instead built a smaller GK104, and release bigger and bigger, step by step, while marketing them like something revolutionary to fire up people to buy the "new" GPUs. I say "new" because I`m pretty sure they have made many ES of several Nvidia GPUs a long time ago, where atleast one design was good enough to be made QS. But instead they wait to see any of the two make a move, then release their QS that was made a long time ago.

Both Nvidia and AMD don`t want to outplay each other, because that lead to an arms race which benefit customers more than the companies. Imagine a beast like GK110/290X from the beginning, mastering all of the games without a hitch. How would they market the new GPUs? Both companies like to use games as examples to get people to buy GPUs.
"This new GPU can play Battlefield 4 in 1600p maxed out"
"Yeah I`ll pass, my GK110/R9 290X released 2 years ago can do that too".
Throw in Mantle and OpenGL optimizations which they pushed out several years ago, that meant that the GPUs is much better utilized with much better performance/hardware (like they do with consoles to squeeze the most out of the hardware), you kinda understand why neither AMD nor Nvidia have bothered making an effort making a better API. Who knows, customers might had only needed a smaller cheaper GPU to play todays games that need a big power sucking GPU that doesnt really perform its best due to inefficient DX API.

I`m not entirely sure how much they pay for silicon from TSMC, although I understand that a GK110 551mm^2 cost more than GK104 295mm^2, a GPU consist of many parts. The silicon itself is just a fraction of the total cost, parts like fan, PCB, vapour chamber, casing etc also cost money. So I have trouble believing that the silicon itself means that the total price have to double (GK104 = $500, GK110 = $1000).

Nether gets to much ahead of the other because of the process that they are both using.

But that said I have a feeling that if NV had released their new GPU's before AMD released their 7000 series we would have seen a much different price/perf picture.

 

[AnandThenMan]

Yes. I've often stewed about something. Why, after all the years AMD and Nvidia, or AMD and Intel have been designing silicon, has not one of the companies ran away with performance? You'd think that over a period of a few decades, one company would vastly overshadow another, not ALWAYS be ridiculously neck and neck especially when AMD for example has next to no funds. They can STILL offer a Titan competitor, when Nvidia has relatively unlimited funds to utterly wipe the floor with AMD. Same with Intel. AMD does not offer a faster CPU than Intels fastest, BUT is STILL able to keep up a few tiers lower with next to no money for R&D and such. When I say next to no money, I'm talking in relative terms compared to Intel or Nvidia.

Seems quite impossible to me and the only explanation that I can wrap my mind around is that they actually are all working together, despite public appearances.
Something isn't right. Hasn't been for many many many years. Only in the last few have I suspected anything like collusion between these companies.

I'm sick (Not really. Just sick of it) over the whole industry. My views have changed pretty radically over the last few years. Anyone ever get this feeling when they see year after year, gen after gen, that one company does not continually walk away from the other? Does not make sense. Even two turtles racing over a 100 mile distance will cross a given finish line at completely different times. Could be a week apart, could be a month, or a year. Best analogy I can come up with on a half a cup of coffee.

Good morning to all, and have a great day!

Interesting conspiracy theory (although hardly new). Maybe AMD is just better at getting the most out of their R&D dollars. Go back 10 years, AMD successfully out thought Intel when they came up with Opteron. Are you saying even back then Intel was colluding with AMD? Makes zero sense given when Athlon first hit, Intel bullied their suppliers to the point where they were afraid to even show AMD products in public, and resorted to putting AMD in white unlabeled boxes (I saw this myself). AMD out thought the industry when it comes to CPU/GPU integration as well, AMD came up with GDDR5 memory etc. etc. the bottom line is when you're the small guy, you have to innovate or die. Established companies very often get complacent and take their foot off the gas. And there is diminishing returns, you can throw 10 billion or 100 billion at a problem and may not see any better results.

Your argument completely falls apart when you apply logic and understand that technology has a natural progression, and the limits of said tech is reliant on factors that all competing companies deal with at any given time. Look at the auto industry, we see the same thing, in general car models are very similar to one another as far as fuel economy, safety, styling cues, technology etc. I suppose you believe the auto industry also has a secret pact?

 

[Erazor51]

Carmacks loathing of AMD, of course.

I'm being serious, Carmack has no love for AMD

Read this:

JOHN CARMACK GIVES MANTLE HIS BLESSING

If the success of the Oculus Rift tells us anything, it's that VR is the way of the future. If it tells is one more thing however, it's that a John Carmack endorsement goes a long way. AMD will be feeling that glow of support today too, as Carmack recently tweeted out how the company's new Mantle API development could be "very helpful," to developers and potentially shake things up in the console scene.

AMD's Mantle system is designed to offer developers an alternative to both DirectX and OpenGL, providing direct access to the GPU's innards and thereby making it theoretically possible for developers to draw several times the power for certain instructions, from the same hardware. This needs to be tried in practice of course, but if Carmack says it's possible, I'm willing to believe it.

However, he did add the caveat that he wouldn't be jumping on the new API any time soon. Ultimately, OpenGL offers very similar performance improvements over DirectX, he says, making it seem a little unnecessary to get aboard the Mantle train just yet.

However, we've heard from EA that its new Frostbite Engine is designed to support Mantle, which means a lot of next-gen games are going to come with that API as an option. It will be really interesting to see what those games perform like on the PC and whether those people with AMD cards end up with a much more aesthetically pleasing gaming experience, than their Nvidia cousins.

Link: http://megagames.com/news/john-carmack-gives-mantle-his-blessing

 

[thilanliyan]

Yes. I've often stewed about something. Why, after all the years AMD and Nvidia, or AMD and Intel have been designing silicon, has not one of the companies ran away with performance?

As far as performance staying close to eachother, I think part of it is that the engineers eventually get to an "optimized" solution to the problem, and whichever direction each company approaches the "optimized" solution from, they eventually get very near each other. I think the only time there are vast differences is when there is a large "rule changes", when unified shaders first came about as an example.

A similar story happens in Formula 1, another highly technical industry. When there are major (technical) rule changes, one or two teams get very close to the "optimized" solution in the first shot so they run away in terms of performance, but over the subsequent years the other teams catch up and there is very little difference in performance. This happened last in 2009 when major rule changes were introduced and the Brawn team ran away with it, but since then several teams have been very close to eachother in terms of performance. This is likely to change again in 2014 when major technical changes are introduced again.

Could be totally off lol, but that's just my take on the closeness in performance.

 

[AnandThenMan]

Even if Mantle ultimately fails, it's already succeeded in shaking things up. :thumbsup:

 

[blastingcap]

You know, if AMD can get into the major engines, especially UnReal, then maybe Mantle has a future after all. But it's vendor-specific so I don't think it will be alone, especially with industry heavyweight Valve allegedly pushing for improved OpenGL to get away from MSFT's Windows monopoly. Personally I am hoping that some non-DirectX alternative arrives ASAP so I am not forced to pay the Microsoft Tax on every PC gaming rig I build.

Read this:

JOHN CARMACK GIVES MANTLE HIS BLESSING

If the success of the Oculus Rift tells us anything, it's that VR is the way of the future. If it tells is one more thing however, it's that a John Carmack endorsement goes a long way. AMD will be feeling that glow of support today too, as Carmack recently tweeted out how the company's new Mantle API development could be "very helpful," to developers and potentially shake things up in the console scene.

AMD's Mantle system is designed to offer developers an alternative to both DirectX and OpenGL, providing direct access to the GPU's innards and thereby making it theoretically possible for developers to draw several times the power for certain instructions, from the same hardware. This needs to be tried in practice of course, but if Carmack says it's possible, I'm willing to believe it.

However, he did add the caveat that he wouldn't be jumping on the new API any time soon. Ultimately, OpenGL offers very similar performance improvements over DirectX, he says, making it seem a little unnecessary to get aboard the Mantle train just yet.

However, we've heard from EA that its new Frostbite Engine is designed to support Mantle, which means a lot of next-gen games are going to come with that API as an option. It will be really interesting to see what those games perform like on the PC and whether those people with AMD cards end up with a much more aesthetically pleasing gaming experience, than their Nvidia cousins.

Link: http://megagames.com/news/john-carmack-gives-mantle-his-blessing

 

[smackababy]

Even if Mantle ultimately fails, it's already succeeded in shaking things up. :thumbsup:

I think what it has a real chance of doing is making MS get off their butt and make DX much better. We know they can (look at the console APIs), they just choose to let it stagnate. They've done the same with IE. It was competitive when it had to be and then after it killed Netscape, it just sat and did nothing for years until Mozilla and eventually Chrome came along and showed people the light of a better browser experience.

Sadly, even if OpenGL and Mantle offer more efficient code paths, developers still are forced to use DX due to the user base of gamers being mostly Windows and the amount of developers familiar with DX greatly outnumbers those familiar with OpenGL or Mantle.

I think OpenGL has a better chance (because it is hardware agnostic unlike Mantle), but I'm sure DX will still be the norm. Even if the big engines get Mantle and OpenGL support (which the big ones already have IIRC), developers have to split time development time between them.

 

[Cloudfire777]

I think what it has a real chance of doing is making MS get off their butt and make DX much better. We know they can (look at the console APIs), they just choose to let it stagnate. They've done the same with IE. It was competitive when it had to be and then after it killed Netscape, it just sat and did nothing for years until Mozilla and eventually Chrome came along and showed people the light of a better browser experience.

Sadly, even if OpenGL and Mantle offer more efficient code paths, developers still are forced to use DX due to the user base of gamers being mostly Windows and the amount of developers familiar with DX greatly outnumbers those familiar with OpenGL or Mantle.

I think OpenGL has a better chance (because it is hardware agnostic unlike Mantle), but I'm sure DX will still be the norm. Even if the big engines get Mantle and OpenGL support (which the big ones already have IIRC), developers have to split time development time between them.

More like whoever gets the edge on the other is the one who can dish out most cash to developers. :ninja:

 

[Techhog]

Now for the Nvidia fanboys to talk about how this will revolutionize the industry while the AMD fanboys talk about how it's no good.

Warning issued for thread crapping.
-- stahlhart

 

[Makaveli]

In 3 weeks nv did what took amd 4 years.
To do so they invented the time machine.

not sure if serious.

Surely NV was working on this far longer than the public announcement of it.

 

[Revolution 11]

Now for the Nvidia fanboys to talk about how this will revolutionize the industry while the AMD fanboys talk about how it's no good.

Well, I am amped for both Mantle and G-sync. Neither will revolutionize the industry, but we will all benefit.

Actually, I am going to go for the plunge here and buy a GPU before December. There is no point in waiting for the real "perfect storm" of competition in the GPU industry which actually is going to appear in 2015-2016.

By 2015/2016, you will have:
1. Mature Mantle and TrueAudio ecosystem

2. Lots of G-sync enabled monitors

3. 20nm GPU cards

4. Increased availability of 4K-enabled games, content, monitors, and GPUs

5. Better developer experience with the next console generation

a. Allowing for a higher minimum standard of ported games and more consistent >4GB RAM use
​

6. DDR4 RAM / PCI-E 4.0 / Sata Express / NVME SSDs Availability

a. Improving RAM and I/O bottlenecks
​

7. More developers realizing that PC games can be more than higher-res console ports

a. Star Citizen, etc
​

What we are seeing right now is just the introductory rumblings of the fight that's to come.

 

[Keysplayr]

Now for the Nvidia fanboys to talk about how this will revolutionize the industry while the AMD fanboys talk about how it's no good.

Is this what you want? I can't see any other reason to mention it. Don't encourage them.

Too late.
-- stahlhart

 

[VulgarDisplay]

Now for the Nvidia fanboys to talk about how this will revolutionize the industry while the AMD fanboys talk about how it's no good.

Honestly this to me is nothing revolutionary. It's just a relaunching of TWIMTBP.

Gsync is something new and exciting though. Mantle is new and exciting also.

 

[smackababy]

Honestly this to me is nothing revolutionary. It's just a relaunching of TWIMTBP.

Gsync is something new and exciting though. Mantle is new and exciting also.

I wouldn't call Mantle new. The optimizations are what developers were doing for DOS before Windows added a layer of obscurity to the hardware and their OS become the standard.

 

[bystander36]

I wouldn't call Mantle new. The optimizations are what developers were doing for DOS before Windows added a layer of obscurity to the hardware and their OS become the standard.

Mantle is more about optimizations in code.

Gsync is something a new way to display frames, and definitely a more optimal way.

 

[Crap Daddy]

Soon we'll need two systems, one running Nvidia and the other AMD, depending on the games we like.

 

[Piroko]

Yes. I've often stewed about something. Why, after all the years AMD and Nvidia, or AMD and Intel have been designing silicon, has not one of the companies ran away with performance? You'd think that over a period of a few decades, one company would vastly overshadow another, not ALWAYS be ridiculously neck and neck especially when AMD for example has next to no funds. They can STILL offer a Titan competitor, when Nvidia has relatively unlimited funds to utterly wipe the floor with AMD. Same with Intel. AMD does not offer a faster CPU than Intels fastest, BUT is STILL able to keep up a few tiers lower with next to no money for R&D and such. When I say next to no money, I'm talking in relative terms compared to Intel or Nvidia.

Seems quite impossible to me and the only explanation that I can wrap my mind around is that they actually are all working together, despite public appearances.
Something isn't right. Hasn't been for many many many years. Only in the last few have I suspected anything like collusion between these companies.

Well, having access to the same patent pool will lead to similar design directions. Reinventing something is much harder than to reinterpret it after all.
Then add, that consumers or other companies (MS, system builders) would prefer for them to aim for the same direction and you end up with a very specific path with narrow goal posts.

You see this a lot currently, like the push for small displacement turbo engines by all but the most daft. Or the push for high DPI, be it sensible or not.

 

[thilanliyan]

You see this a lot currently, like the push for small displacement turbo engines by all but the most daft.

Sad days, but the way of the future, and turbos do have some good characteristics. BUT, give me a screaming naturally aspirated engine for my "Sunday" driver.

 

[Cloudfire777]

Tim Sweeney from Epic Games and John Carmack from ID Software - Sceptical to Mantle but want more development on OpenGL

02:47PM EDT - What do you think of AMD's Mantle?

02:48PM EDT - Carmack: My opinion is, in terms of action items, should NV have a response to Mantle? Unequivocally no, a horrible mistake if NV got panicked by this and made some lower level API, you already have good low level access through NV GPU extensions, give you as much performance as you want if you're fine with proprietary extensions

02:49PM EDT - Carmack: Mantle only became interesting because of their dual console wins, the landscape does matter that they have both major console wins with similar architectures, not a stupid thing that AMD is doing at this point, could have some implications for Steam, if MS and Sony embraced it that would be very powerful for AMD but it doesn't look like they're going to (at least MS), if I was still doing all of the major tech coding I probably would not be embracing Mantle right now but there would be days where it would be extremely tempting

02:51PM EDT - Sweeney: there's some good ideas in Mantle we really liked the idea of having low overhead access to the GPU, if you look back at DX/OGL, there's a lot of overhead in those APIs and the fact that they date back to the old SGI model of rendering which is very different than the current model, potentially unified memory, good ideas there I hope it really helps the OGL community and MS evolve their APIs. Would I like to have a low level API? Yes. Five of them? Absolutely not. To have yet another API on the PC that's still different than the DX API and OpenGL, and OpenGL ES and different than PS4 low level API, I don't think it's a good idea.

http://www.anandtech.com/show/7437/tim-sweeney-johan-andersson-john-carmack-nv-montreal-live-blog

 

[Imouto]

Why you keep posting whatever those guys say when they're on Nv's payroll?

 

[SiliconWars]

This seems to be Nvidia's way of responding to Mantle, by getting Carmack and others to downplay it as much as possible.

 

[bystander36]

You have to recognize the truth in their words, even if you don't want to believe it. I know you want Mantle to be the end all of gaming, but the reality is it will take extra work for it to happen. Work the dev's aren't likely going to jump on board so quickly.

I personally am not excited about adding another API to the mix, but I am curious to see how it works. The only good thing I can see come out of this, is it'll push MS to improve DX, or maybe OpenGL gets improved and DX and Mantle both go away.

More API's is not the goal. Consolidation is a better goal.

 

[Grooveriding]

This seems to be Nvidia's way of responding to Mantle, by getting Carmack and others to downplay it as much as possible.

I don't know about that. Carmack seems really opinionated and fairly passionate about his views. Sweeney on the other hand spent a large portion of his time plain shilling for nvidia with the constant referencing to gsync and praising this, that and the next thing.

That Q&A at the end was the interesting part of the whole thing. I think the young lad held his own against Sweeney/Carmack's attempted onslaught pretty well.

 

[Kenmitch]

Tim Sweeney from Epic Games and John Carmack from ID Software - Sceptical to Mantle but want more development on OpenGL


http://www.anandtech.com/show/7437/tim-sweeney-johan-andersson-john-carmack-nv-montreal-live-blog

But every day they too are surfing the web for more info on Mantle....Go figure.

Would be interesting to see there comments once the BF4 Mantle update drops....Maybe AMD and Dice should have a special event!