The game console living room fortress will not be breached by cellphones & tablets.

Tegra K1 is 192 Cuda Cores right? Even lowly X1 is 768 shaders. ~4X.

But K1 in an actual tablet or phone is probably going to be clocked pretty low for heat, and have crap bandwidth, so the real differential will probably be 10X+.

But that K1 might deliver X360+ graphics, which frankly is enough to already kinda sorta ape X1/PS4 class if you squint (just look at stuff like Halo 4 or Crysis 3). And you will be squinting looking at your 5" phone screen anyway.

But, I'd argue no developer will spend the $ anyway on a "triple A" mobile game, for the most part. Great visuals are a function of dollars spent and manpower as much as hardware. Even if mobile gets the latter, they'll lack the former.

It really doesn't matter though imo, because touch devices dont have buttons. Even more than graphics it's imo what really prevents tablets/phones from threatening the core gaming market.

Anyways K1 is really nothing special I assume, but Nvidia hype. The latest Qualcomm and Power VR will probably equal or exceed it, without the hype. I mean they talked a lot of smack about EG, Tegra 4 too, and those didn't particularly lead the class.
 
- LOOK AT THE FIGURES FOR CELL IN THE CPU COMPARISON! HAHAHAHA!!

It's funny but if you look at the note at the bottom of the slide you can see that that for the CELL - they don't include the SPU's or any of it's GPU support.
 
Except you are not going to get an increase of 6.5x in performance per watt. You'll be lucky to get 3X, IMO. Even TSMC only claims 25% power reduction from 28 to 20nm.

My explanation for the 6.5x were explained. This actually a very conservative number because I didn't even consider the architectural changes that have been leading to much better theoretical-to-real life approximations.
Where does your 3X number come from?

TSMC claims 25% power reduction between mature 28nm and fresh 20nm, which is really conventional, actually.


Maybe not about their own - where they are probably just 'optimistic' - but Nvidia are as full of shit as every other company.

Look at the blown up version of the image of the console 'comparison' half way down this article:

http://www.pcworld.com/article/2084...generation-mobile-processor-the-tegra-k1.html

A cursory check:

- Xbox 360 is way beyond DX9
- Xbox 360 is under 100W for the entire system, at the wall, powering a billion X DVD drive, 15W or so of Kinect, a mechanical HDD, wifi, memory etc, on a combination of 45 and 55nm
- Nividia are comparing a chip to an entire system
- LOOK AT THE FIGURES FOR CELL IN THE CPU COMPARISON! HAHAHAHA!!

If you look at the pics, you'll see they're comparing to the CPU+GPU combination of the first generation of X360 and PS3 that came out in 2005/2006.
If they were comparing system vs. chip, then the X360 would be 200W and the PS3 would be 230W.
Xenon is closer to DX9 than anything else.


In conclusion: Nvidia are as full of shit as any other company.

Tablets won't eclipse the 100+ Watt 4Bone any time soon, and certainly not by 2017. We'll be lucky if we've moved beyond 20nm for mainstream products by 2017. New node ramp got cramp.

Yes, nVidia's marketing department is as full of shit as any other marketing department.

However, that doesn't change my thinking that we'll have something at least very close to a xbone/ps4 in tablet form by mid/end 2017.
In 3.5 years? I have no doubts.
 
My explanation for the 6.5x were explained. This actually a very conservative number because I didn't even consider the architectural changes that have been leading to much better theoretical-to-real life approximations.
Where does your 3X number come from?

TSMC claims 25% power reduction between mature 28nm and fresh 20nm, which is really conventional, actually.




If you look at the pics, you'll see they're comparing to the CPU+GPU combination of the first generation of X360 and PS3 that came out in 2005/2006.
If they were comparing system vs. chip, then the X360 would be 200W and the PS3 would be 230W.
Xenon is closer to DX9 than anything else.




Yes, nVidia's marketing department is as full of shit as any other marketing department.

However, that doesn't change my thinking that we'll have something at least very close to a xbone/ps4 in tablet form by mid/end 2017.
In 3.5 years? I have no doubts.

Really I dont see any change from previous generations. Remember when PSP was released? It delivered a PS2 like experience in a small screen years before the PS3 came along. The PSVita came like when? End of 2012? Just like in the past only recently are we getting near last gen quality visuals, but certainly not near next gen quality visuals on a tablet this soon
 
3 node shrinks is:
1: 20nm
2: 14nm
This is effectively one node transition in terms of density+power improvement for the foundries. The power scaling of 28nm to 20nm is poor. The density scaling from 20nm to the hybrid 16/14nm nodes is small.

The time to an economically justifiable 10nm transition for mobile is a question mark or three.
 
But that K1 might deliver X360+ graphics, which frankly is enough to already kinda sorta ape X1/PS4 class if you squint (just look at stuff like Halo 4 or Crysis 3). And you will be squinting looking at your 5" phone screen anyway.

That right there is the key to this whole thing that most seem to miss. You will be able to provide an "identical" game experience on a phone because on such a small screen you don't need to render all the same details as a console does on a 60" screen. The problem currently was there was still too much of a power gulf between phone and console, but that is in the process of being solved to where soon it won't matter and/or many won't see enough difference to care. In other words a 5" screen will be able to provide a visually identical experience to a 60" screen even at much lesser wattage. Tablets have bigger screen but also have bigger batteries so they can go higher wattage wattage there to account for that screen jump again achieving the same goal of having a 10" screen provide a similar visual experience to a 65" screen. The difference between all these devices visually will begin to blur because you don't need the same amount of horsepower to achieve a similar visual goal on a smaller displays.


But, I'd argue no developer will spend the $ anyway on a "triple A" mobile game, for the most part. Great visuals are a function of dollars spent and manpower as much as hardware. Even if mobile gets the latter, they'll lack the former.

That's the beauty of it all because they won't have to. In the case of k1 it's Kepler, the same tech as the desktop parts, and Unreal already runs on it. There is no re-learning, no re-architecting, none of the above. They don't have to spend $$$ on the software anymore, or at some point won't have to anyways. Plus the hardware now is all by default being built as mobile parts first, then upscaled for desktop and other use. Once you have such a simple setup there becomes less reason to not port your games everywhere.


It really doesn't matter though imo, because touch devices dont have buttons. Even more than graphics it's imo what really prevents tablets/phones from threatening the core gaming market.

They all have gamepad support by now no?


Anyways K1 is really nothing special I assume, but Nvidia hype. The latest Qualcomm and Power VR will probably equal or exceed it, without the hype. I mean they talked a lot of smack about EG, Tegra 4 too, and those didn't particularly lead the class.

Yeah it will be exceeded soon, that's just makes the case all that much stronger. Think of the goal, buy once use anywhere. I already do that right now on all my devices except phones which are still lagging, but a game I buy I can play on my tv if I want, play the same game on my laptop if I want or my tablet if I want, anytime anywhere in any location. I already do that today. That alone has made consoles horribly expensive and unattractive to me with their proprietary hardware, proprietary software and locked down ecosystem that doesn't play nice with anything else I own. Console gaming offers a worse visual experience than my desktop pc and none of the flexibility of letting me play my own purchased software on the devices I own. That operating model will die a painful and much deserved death in the future especially once phones can start running that same software that I already use on my other devices. This will make the prospect of spending more money to buy a game only console that forces you to re-buy everything and not let you use any of it anywhere else a dead proposition. Microsoft seems to be planning for that albeit at a glacial pace, but they at lease seem to have it on their radar to join the inevitable merge of it all.
 
Tablets and phones are not consoles, It also works the other way around. I don't think the specs and graphics are that relevant to the topic, simply put it is not a ceteris paribus comparison the same game doesn't look the same on a + 35inch screen seen from a sane distance than on a screen between 5 and 10 inch you hold in your hands. It is a completely different experience.
Phones and tablets could steal some of true most casual core gamers away but I see the whole thing as positive, it is an opportunity for publishers to leverage older IP, pretty much consoles will subsidize for the game developments of mobile games sold for cheap.

I think android stb with sane graphics capability could be more threatening to the Console, though it is like opening a Pandora box, business is unproven and it is unclear how AAA games would be fund, how publishers margins would be affected, etc. Ultimately nobody wins a race to the bottom though there are always willful contenders...
 
I've never seen a game that looks as good on a trillion x trillion 1 inch retin0r display as it does on a 42" 1024 x 768 plasma. And triggerless / shoulder-buttonless controls mapped to touch screen blow chunks.

Even if the hardware is shrunk down to the size of a wasp, there will always be a need for something that plugs into a big tv and has a controller.
 
Even if the hardware is shrunk down to the size of a wasp, there will always be a need for something that plugs into a big tv and has a controller.

For sure, I do that already. I plug in my Mac Air to a plasma tv in a different room, sit down on the couch and play all my existing games with my 360 controller. When I'm on the road or away I play the same games on the same laptop. When the main tv room is free I play the same games on that tv on a different machine. The mantra is the same, always playing the same content that I've already paid for in whatever place I happen to be on whatever device I have with me. That functionality already exists today. Yes it's common place in the console world to lock all your content to one tv, to not let you play it on the road, to make you throw it all away and have to rebuy it when the next console comes out, to make you buy more proprietary hardware to play your existing software, etc, but that's just the console worlds archaic business model that is ultimately doomed. The limitations thus far have been that tablets and phones have still been on relatively weak hardware and while tablets allow you to use all your same purchase and software anytime anywhere, phones still don't fully allow that. With the fast rate of mobile hardware advancement though that is all set to change in the near future not only solving the problem of mobile hardware that are also capable gaming machines, but solving the business problem of making it much easier for the software makers to port their software to all such devices without having to re-write it all from scratch.
 
For sure, I do that already. I plug in my Mac Air to a plasma tv in a different room, sit down on the couch and play all my existing games with my 360 controller. When I'm on the road or away I play the same games on the same laptop. When the main tv room is free I play the same games on that tv on a different machine. The mantra is the same, always playing the same content that I've already paid for in whatever place I happen to be on whatever device I have with me. That functionality already exists today. Yes it's common place in the console world to lock all your content to one tv, to not let you play it on the road, to make you throw it all away and have to rebuy it when the next console comes out, to make you buy more proprietary hardware to play your existing software, etc, but that's just the console worlds archaic business model that is ultimately doomed. The limitations thus far have been that tablets and phones have still been on relatively weak hardware and while tablets allow you to use all your same purchase and software anytime anywhere, phones still don't fully allow that. With the fast rate of mobile hardware advancement though that is all set to change in the near future not only solving the problem of mobile hardware that are also capable gaming machines, but solving the business problem of making it much easier for the software makers to port their software to all such devices without having to re-write it all from scratch.

doesnt work with

that right there is the key to this whole thing that most seem to miss. You will be able to provide an "identical" game experience on a phone because on such a small screen you don't need to render all the same details as a console does on a 60" screen.

Because now you do need to provide that detail.
 
Because now you do need to provide that detail.

No, you would use a stronger box for that if the best possible visuals is what you demand. Right now you can't do that with consoles because they make you re-buy everything. You have to buy console specific software that works no where else but on that one console box, which in 2014 is a silly concept. And after paying all these console taxes you aren't even left playing the best version of the software, you are stuck playing it on mobile based console hardware on your large screen tv which will run with crappier visuals and crappier framerate than it could be.

Instead, for those that care about getting the best visuals, use a desktop pc to play games on a big tv when you want full quality and best framerate, use laptop hooked to a tv when it's convenient or if you don't care as much about getting the best visuals available, and use your tablet or phone when on the go, on your hammock, at a friends place or whatever. No console taxes necessary, use your same digital software everywhere. History has shown that people will pick convenience over quality, and a fully integrated software setup like I described would ultimately win out on convenience and you can still get the best quality available if you want it as well with more powerful hardware hooked to your tv, for those so inclined. I suspect many would forgo the more powerful hardware anyways and be content with playing whatever visuals their tablet can throw to the tv. Right now console gives you none of the above. You get no software backward compatibility so you have to throw all your purchases away when you get new hardware, you can't get the best visuals, you don't get freedom to use your software on other hardware, you can't use your software on a mobile on a plane if you wanted, you are locked forever to one manufacturer, etc, and you get to pay more for all those limits and restrictions. This business model simply won't last.

Right now the handicap is mobile gpu's which are too weak thanks to a lack of focus in that realm. As we are seeing, that is changing at a very rapid pace to where my next ultrabook will give me 360 quality visuals on the go, or to a tv if I want it that way. Soon after my phone will do the same, and so on and so forth. The existing console model is basically being eaten up at both ends. High end visuals keep getting better and better, so those demanding the best will leave consoles as they get more and more dated. On the low end mobile gpu's get stronger and strong so the yearly cycle hardware like tablets, etc will munch away console customers as they become satisfied enough with what tablets, phones or ultrabooks can do. In the middle there are those fed up with having to throw away their software all the time and/or having it locked to only run on the one console, and over time they will walk away from consoles and go to less restrictive hardware models. Finally there's those that already own a tablet, phone or whatever who were perhaps console buyers in the past but now their other devices provide enough entertainment to where they no longer will buy a console, at least consoles as we currently know them.
 
For sure, I do that already. I plug in my Mac Air to a plasma tv in a different room, sit down on the couch and play all my existing games with my 360 controller. When I'm on the road or away I play the same games on the same laptop. When the main tv room is free I play the same games on that tv on a different machine. The mantra is the same, always playing the same content that I've already paid for in whatever place I happen to be on whatever device I have with me. That functionality already exists today. Yes it's common place in the console world to lock all your content to one tv, to not let you play it on the road, to make you throw it all away and have to rebuy it when the next console comes out, to make you buy more proprietary hardware to play your existing software, etc, but that's just the console worlds archaic business model that is ultimately doomed. The limitations thus far have been that tablets and phones have still been on relatively weak hardware and while tablets allow you to use all your same purchase and software anytime anywhere, phones still don't fully allow that. With the fast rate of mobile hardware advancement though that is all set to change in the near future not only solving the problem of mobile hardware that are also capable gaming machines, but solving the business problem of making it much easier for the software makers to port their software to all such devices without having to re-write it all from scratch.

I'd rather buy a new machine every 3 - 5 years then buy a new smartphone every year. If hardware is going to advance so much that by 2017 the graphics fidelity of PS4/XB1 is matched by mobile devices then what prevents Sony or MS from releasing a new console that's another generational leap above the one we just got? MS is surely able to do it in terms of XB1 platform what with being x86 and AMD APU and all. All they need to do is transfer the current ecosystem to the new platform and all new games released will be playable by both devices. The argument that tablets/smartphones will advance to the current systems is cool and all but if I can have that type of performance in a handheld, then that means I can have a shit ton more in some actual plug in hardware. I'm not sure how this destroys the console model, maybe the companies will just need to adapt by releasing successors at a faster rate. The hardware could be sold for a profit and software compatible with all versions of the console.

Again, I'd rather spend $400 to $500 on a new home machine every few years than buying a new tablet/smartphone at similar prices, if not more, at a much faster rate.
 
I'd rather buy a new machine every 3 - 5 years then buy a new smartphone every year. If hardware is going to advance so much that by 2017 the graphics fidelity of PS4/XB1 is matched by mobile devices then what prevents Sony or MS from releasing a new console that's another generational leap above the one we just got?

Yeah I agree, I'm talking about the current console model as we know it being basically dead. That of 6-8 year cycles of the same hardware, all software locked down to one device, no backward compatibility, no device freedom, etc, that will all go the way of the dodo. MS to me seems to be positioning themselves to where they could introduce new hardware every 3 years if they wanted to, and additionally they could make their new console fully compatible with their phones and tablets. That is the direction consoles will have to go in my mind and it looks like Microsoft is going in that direction. Ultimately the xbox console will become just another app playing device, it will no longer be it's own locked down universe but one of many devices that plays along with others, shares apps with others, etc.
 
Yeah I agree, I'm talking about the current console model as we know it being basically dead. That of 6-8 year cycles of the same hardware, all software locked down to one device, no backward compatibility, no device freedom, etc, that will all go the way of the dodo. MS to me seems to be positioning themselves to where they could introduce new hardware every 3 years if they wanted to, and additionally they could make their new console fully compatible with their phones and tablets. That is the direction consoles will have to go in my mind and it looks like Microsoft is going in that direction. Ultimately the xbox console will become just another app playing device, it will no longer be it's own locked down universe but one of many devices that plays along with others, shares apps with others, etc.

I'm in full agreement here. I think at this point in time it should be a primary focus of both companies in getting their ecosystems fully fleshed out to where their ecosystem ends up becoming the base platform versus a specific machine. This is an area where MS is at a much greater advantage currently. I agree with you in seeing them position themselves for this. That QoS patent MS had about a scalable architecture has made me wonder if MS may be working on bringing a new machine to market sooner than later. I'm comfortable with a new release every three years if I can just buy the new system and login and wait forever how long to download and play my games at higher settings. It's how I imagine a Steam console would work as well, with a new version coming out every few years.
 
More important than theoretical graphics, no one is going to invest tens of millions of dollars into making the mobile equivalent of GTAV when Angry Birds and Candy Crush rake in their millions because people are stupid and buy "1000 coins for $10!". That math will *never* add up.

Or put another way, the graphics of Infinity Blade III are amazing, but the gameplay is quite dull. Compare that to any top tier PS2 game and I don't see why you'd ever pick what's available on the iPad to play for longer than 10 minutes. And the GPU performance on the iPad Air is several orders of magnitude better :rolleyes:
 
I don't see why you'd ever pick what's available on the iPad to play for longer than 10 minutes. And the GPU performance on the iPad Air is several orders of magnitude better :rolleyes:

There are lots of great deep games in iOS. I've sunk more hours in Galaxy on Fire 2 than the vast majority of my PS3 library.
 
More important than theoretical graphics, no one is going to invest tens of millions of dollars into making the mobile equivalent of GTAV when Angry Birds and Candy Crush rake in their millions because people are stupid and buy "1000 coins for $10!". That math will *never* add up.

Or put another way, the graphics of Infinity Blade III are amazing, but the gameplay is quite dull. Compare that to any top tier PS2 game and I don't see why you'd ever pick what's available on the iPad to play for longer than 10 minutes. And the GPU performance on the iPad Air is several orders of magnitude better :rolleyes:

Part of that is due to the controls, I think now that controller standards are available for both Apple and Google phone and tablets things can improve. We're also starting to see development tools which are targeting these platforms in meaningful ways which could help with development cost and time to market.

That said I think the memory limitations are actually one of the biggest issues that can't be solved very easily, we're already starting to see games in the tens of GB for new consoles and even last generation games were several GB in size this isn't to say a game would need to be as large on a tablet or phone but a true game would be substantially bigger than what many of these titles are today.

Perhaps these platforms will evolve to be used to stream content similar to Playstation Now appears to work. That makes a lot of sense bc the developer can make one version of the game which works on Android and Apple devices that is truly platform agnostic but of course that sort of streaming service for now is more realistic when users are at home and can piggyback off the home network and use a bluetooth controller without having to carry it around in a pocket. When gamers are truly out and about the challenges of virtual controls will likely remain an important challenge to work around.

IMO I think we'll see dumbed down ports of AAA titles both current releases and stuff that was successful on older console platforms, or alternative versions of titles... EA could probably do a decent version of Madden for phones and tables that focused more on picking the plays, trading players and fantasy football - kinda surprised nobody has attempted to make a decent football manager game yet actually....

The other issue however is pricing, mobile and tablet consumers haven't proven to be willing to pay what console owners are for entertainment but the universe of potential consumers is much larger and used games don't cannibalize potential sales for developers.
 
Trying to chop up a AAA game piecemeal into something that can easily be nickel and dimed isn't going to translate well into mobile and I pray dies a swift death on consoles. Micro transactions are total shit.
 
Do we have anything today that can deliver the same power as the 360/PS3 in a Tablet/Phone?

And just the size of the games of the PS3 area seems would be an issue today, yes there is the odd game out there that is very big, but from my experience you reach the limit very fast with big games when you can't swap discs and doesn't have a 500GB hard drive in your phone.

In any case, if we can get PS4 power in a tablet in 2017 that handles games between 25-50GB we might get lucky and get new consoles earlier in this round :)
 
Back
Top