NVIDIA Maxwell Speculation Thread

[Grall]

( why not seen compatibility between iOS games and OSX ? )

I assume, because OSX does not have touch support to the level iOS does...

Also, Apple might actually want to keep the two separate...for whatever reason(s) they may have.

 

[lanek]

I assume, because OSX does not have touch support to the level iOS does...

Also, Apple might actually want to keep the two separate...for whatever reason(s) they may have.

I think this is a minor problem, most "touch games" are just try to reproduce what other physical input do .. Games with pure touch only, specific input, are really rare on smartphone, tablet ..Most just try to emulate physical ( for most gamepad ) use..

If Nvidia can sell you a console with Android games, i dont see why a iOS compatibility with their own "desktop" system cant work... Marketing wise, they could try a big thing there... how it work in reality ? not much that Android games can surviive in front of PC gaming.

 

[silent_guy]

Im pretty sure it will... But what will be the point, at least for gaming and softwares who are not available only on OSX ? ... Dont you think it will be better for a developper to develop / port a game to "OpenGL Vulkan", who will work with the hardware used by Apple ( AMD or Nvidia ).. than again port it to Apple Metal ? ... ( i mostly imagine a case where game is developped with DX12, but ported to Vulkan then, and then will need again ported to Metal OSX.. )

I understand the hype for Vulkan, but I'm not convinced it will have the big impact that everybody else seems to be believe it will have unless the two elephants in the room declare their support for it: Apple and Google.

For the PC market, Vulkan doesn't really add material benefits over DX12, at best some compute related things that Microsoft could add if they wanted to. This market has always been DX and I don't see a reason why it'd switch to something else anytime soon. The argument that was always used in favor of Mantle, "it's easy to make an existing engine like UE4 to support both", works the other way around as well in favor of DX12. As usual, we can expect DX12 to have a way better test suite than Vulkan will ever have, so in terms of compatibility, DX12 will still have the upper hand.

And so, Vulkan will be an OpenGL replacement only.

Professional OpenGL applications are not exactly the most dynamic field in existence, so that's going to be a long term thing at best.
iOS has Metal.
Linux is still very much niche and will probably stay that way. (I fail to see the appeal of a Steambox over an identically equipped Windows machine.)
I put my money on Metal for OSX rather than Vulkan for OSX.
The only thing left is Android. Vulkan has a chance for larger adoption once Google commits to that. But until then, I don't see it going anywhere soon.

 

[AlexV]

Interestingly enough, there was no place left for the Apple and Google logos on that Vulkan supporter/announcement/thingamebob. They're clearly not as important as Canonical or Codeplay, but you'd think that some concessions could be made!

 

[Alexko]

I can see why Apple might not want to support Vulkan, since they already have Metal and they like to do their own thing, mostly so they can control everything.

But I don't understand why Google isn't supporting Vulkan. After all, the more successful Vulkan is, the more viable Linux-based OSes (Android and Chrome OS included) become as gaming alternatives to Windows. Maybe they plan to support it but prefer to keep quiet until they have something to show off.

 

[AlexV]

But I don't understand why Google isn't supporting Vulkan. After all, the more successful Vulkan is, the more viable Linux-based OSes (Android and Chrome OS included) become as gaming alternatives to Windows. Maybe they plan to support it but prefer to keep quiet until they have something to show off.

Why would Google care about e.g. Ubuntu being better for Crysis 4? Android and Chrome OS are Google OSes, Google would care about those, and perceived osmotic benefits derived from the fact that at some point they were Linux are likely to be overstated. So the question is if Vulkan is something that would bring tangible benefits to those two products.

 

[Silent_Buddha]

Why would Google care about e.g. Ubuntu being better for Crysis 4? Android and Chrome OS are Google OSes, Google would care about those, and perceived osmotic benefits derived from the fact that at some point they were Linux are likely to be overstated. So the question is if Vulkan is something that would bring tangible benefits to those two products.

As long as they remain the dominant player (unit marketshare) for smartphones, they likely won't care. The dominant player rarely cares about what smaller players do unless they are perceived to be a threat (DX - Windows v. OGL - everyone else for example). Apple already has Metal so they don't necessarily need a low level API and Android is still dominating the market (so Google won't care). The only thing that might get Google to do something is if Windows phones start to take away Android smartphone marketshare due to more performant games (how performant do they need to be on smartphones?) or significantly better battery life playing the same games (if game is already performant enough why not save battery power instead?).

But that's a bit of a stretch to think that Windows Phone will be threatening Android's grasp of the market anytime soon, if ever. So yeah, there's not a huge incentive for Google. Although there is definitely a potential for Windows tablets to eventually overtake Android tablets. So, who knows.

Regards,
SB

 

[silent_guy]

Google probably first needs buy-in from a number of SOC makers before it can announce its support.

 

[rpg.314]

Anyone wants to set up a betting pool about a Metal for OSX announcement at the next WWDC?

I'll take you up on that.

My bet, Vulkan for os x at wwdc 2015.

Why? Mac is not a priority platform. A simpler API with most of the heavy lifting done by others is best for apple. They would ofcourse be controlling the future of ios by keeping metal around.

 

[silent_guy]

I'll take you up on that.

My bet, Vulkan for os x at wwdc 2015.

Excellent! There's a virtual beer at stake here!

 

[Alexko]

Why would Google care about e.g. Ubuntu being better for Crysis 4? Android and Chrome OS are Google OSes, Google would care about those, and perceived osmotic benefits derived from the fact that at some point they were Linux are likely to be overstated. So the question is if Vulkan is something that would bring tangible benefits to those two products.

But wouldn't it be easier to make Android/Chrome OS more appealing by supporting Vulkan rather than by introducing their own low-level API, which would require quite a bit of work and be unlikely to have any tangible benefits over Vulkan?

 

[liquidboy]

- Deleted by the poster as this post has nothing to do with this thread -

 

[willardjuice]

But wouldn't it be easier to make Android/Chrome OS more appealing by supporting Vulkan rather than by introducing their own low-level API, which would require quite a bit of work and be unlikely to have any tangible benefits over Vulkan?

I'd argue not in the long run though. Google would gain full control of the API. Do you think IHVs wouldn't support whatever custom API Google comes up with? Why should Google cede any power?

 

[rpg.314]

Excellent! There's a virtual beer at stake here!

let's make that vulkan for os x, whenever vulcan is ready.

It's pointless to bet virtual beer over schedules into which I have no insight.

 

[Deleted member 2197]

Moderators .... Pls clean-up thread.

 

[Albuquerque]

I'm just here for the 980ti. Wake me when it arrives

 

[Deleted member 2197]

GTX Titan X benchmarks ???

The cards have been tested properly with a Core i7-5960X eight-core processor, and the scores are both single-GPU and 4-way SLI with on 3DMark 11, with its "extreme" (X) preset. The card scored X7994 points in a single GPU run and a whopping X24064 points in 4-way SLI. Again that would be the test in EXTREME mode. The screenshots that leaked through Videocardz confirm the announced 12GB memory partition, but it also is listing a boost core clock of 1222 MHz and a memory clock of 1863 MHz (x4) for the Quad SLI setup. We do assume here that the card would be overclocked or tweaked a little, but these clocks are yummie for an 8 billion transistor encounting product!

Clock wise the single GTX Titan X would be clocked around 1.0 GHz as the screenshot indicates. The memory is clocked at 1753 Mhz (x4) = 7.0 Ghz effective. And these numbers do make more sense. The GTX Titan X is expected to get 3072 shader processors, 192 TMUs, 96 ROPs, and a 384-bit wide GDDR5 memory bus with 12 GB of memory. The majority of specs are based on the GM200 Quadro counter-part. What remains weird though is that the entry is listed as 'Generic VGA', without a driver you simply can not activate SLI or even the single GPU, so it really should state something 'GeForce'.

http://www.guru3d.com/news-story/geforce-gtx-titan-x-3dmark-benchmarks-surface.html

 

[Frenetic Pony]

GTX Titan X benchmarks ???



http://www.guru3d.com/news-story/geforce-gtx-titan-x-3dmark-benchmarks-surface.html

Seems a guaranteed fake. A GTX 980 also runs at up to 1.2ghz, indicating (here) there's no clock advantage for the Titan. Meanwhile the new Titan manages to almost double the score of a 980 despite only a sixty percent increase in transistor count. So, somehow, on the same architecture and with the same clock speed, the new Titan scales 1.5x in performance with transistor count over the 980.

 

[lanek]

http://wccftech.com/nvidia-geforce-...erformance-uplift-gtx-980-234way-sli-results/

 

[silent_guy]

If everything is scaled up by 50%, and the clocks are a little lower, you don't even need (potentially fake) benchmarks have high confidence in it being 40-45% faster than a GTX980.