Fact: Nintendo to release HD console + controllers with built-in screen late 2012

More speculation about the controller. I'm still not impressed by the idea of a gamepad with a touch screen. Even if it has hepatic technology and it works amazingly well, all you get is a lot of buttons with pretty pictures on them. How is that supposed to be the next step in gaming?

If Nintendo wants to build on the success of the Wii, the next step would be hand gesture control. Imagine a device you wear on a forearm with cameras directed at your hand. With the help of tilt and motion sensors, the setup should allow you to reconstruct an accurate model of your entire arm in 3D space--down to the finger tips. That opens up tons of possibilities. To cast a spell in a RPG game, you just do a casting motion. To employ telekinesis, just cup your hand and direct it at the object you want to grab. To throw a split-finger fastball instead of a two-seamer in a baseball game, just change your grip. Such a controller would probably have a touch screen for things like play-calling. The sci-fi cool factor along warrant its inclusion.

Kinect already does 95% of the things you describe. It wouldn't be new....
 
I would imagine that anything new would take cues from recent PowerPC developments. Xenon is pretty simple, simpler than even Broadway in some ways (OOE for ex).


Thanks for pointing out that Broadway has OOE. I was going to ask.

It's interesting that they stress both easy development and easy porting from PC and 360. That leads me to believe that it will be either similar to PC or similar to 360. So maybe Llano is an option? I doubt they can use anything closely resembling 360 architecture without MS being unhappy.

Perhaps 3 greatly simpified POWER7 cores ?
 
On the CPU, easy of use by experienced programers and for multi platforms games is (should be) one of the prioritys for Nintendo, that means two things IMO, 1) or it is based on Xenon (and must code can be used and even run better) or it is like a X86 CPU (and must code can be used just need some fine tunning console style) 2) it will perform as good as Xenon without any hassle (and that should probably be easy to be profitable in a 200-250$ console).

If that means Xenon based CPU or a liano/fusion based CPU, I will vote in whatever make a better deal in a mid-long term.


Kinect already does 95% of the things you describe. It wouldn't be new....

The ability of playing/Shoot a FPS on Kinetic alone would be a mini revolution IMO, at least for core gamers :D

Anyway I really like of the above idea, and I am found off the idea of a good collection of a few things that as a whole are indeed a innovation.

Really a touch screen as sole innovation is so pre DS and pre IPAD...
 
Kinect already does 95% of the things you describe. It wouldn't be new....

That's like saying you can do 95% of the things you can do with your hands while wearing boxing gloves.

Accurate tracking of finger positions definitely is the next logical step. I don't think the Kinect approach will ever give you enough precision. It has to be a glove or cameras mounted near the hands.

hepatic technology :|

LOL. I typed haptic at first, but Chrome said it was misspelled :???:
 
Here's a video showing the kind of motion capture you can do without using markers:

http://www.youtube.com/watch?v=dTisU4dibSc

Obviously, this technology can't be used in a consumer product for full-body scanning. No one's living-room is going to be a dark empty stage and we can't expect a consumer to go calibrate a bunch of cameras. But if we focus only on the hand though, then maybe it's workable. The human hand is a known shape. A pair of cameras at the wrist and another pair further back should capture enough information to let you reconstruct a fairly precise 3D model.
 
I doubt they can use anything closely resembling 360 architecture without MS being unhappy.

This is an interesting facet, wasn't the story that MS essentially "stole" much of the IBM/Sony PPU design from Cell for their tri-core?

I'm wondering if, the IP is set up in such a way that "stealing" this tech is perfectly legal, Since nothing was done legally about MS's alleged theft (perhaps IBM has final control of the patents to whore out as they wish?).

And obviously it makes a difference to our speculation expedition whether we think Nintendo can just copy (for the most part) the MS tri-core design, or must use something fundamentally different. I rather suspect the former.
 
Obviously, this technology can't be used in a consumer product for full-body scanning. No one's living-room is going to be a dark empty stage and we can't expect a consumer to go calibrate a bunch of cameras. But if we focus only on the hand though, then maybe it's workable. The human hand is a known shape. A pair of cameras at the wrist and another pair further back should capture enough information to let you reconstruct a fairly precise 3D model.
4 cameras?! A next-gen Kinect depth-camera would be the solution. Lower cost, works with multiple users, and isn't dorky. Strap on cameras on the arm that only allow arm tracking is overkill. Unless you need pixel-perfect resolution to recreate something like playing a piano, a camera set back can give enough detail for major hand changes (open/closed) which'll be enough for pretty much every game, I'm sure.

Regardless though, there's no reports of Nintendo do any such thing as arm tracking, so it's a pipe-dream at the moment. And I reckon MS will have pretty good arm-tracking in their next Kinect implementation whenever that appears.
 
4 cameras?! A next-gen Kinect depth-camera would be the solution. Lower cost, works with multiple users, and isn't dorky. Strap on cameras on the arm that only allow arm tracking is overkill. Unless you need pixel-perfect resolution to recreate something like playing a piano, a camera set back can give enough detail for major hand changes (open/closed) which'll be enough for pretty much every game, I'm sure.

The 3DS has three cameras, so four isn't crazy. And they can be monochromatic. Probably add like $5 to the BOM. If you want to do finger tracking in a mass consumer product, cameras are the only economical way. A strap-on controller is dorky, I admit, but you need gyroscopes and accelerometers to track arm movements with high precision and the user can't be holding a ward in this scenario.

Piano is too dull and stuffy for a video game, but a guitar playing game controlled by finger movements could be popular.
 
This is an interesting facet, wasn't the story that MS essentially "stole" much of the IBM/Sony PPU design from Cell for their tri-core?

I'm wondering if, the IP is set up in such a way that "stealing" this tech is perfectly legal, Since nothing was done legally about MS's alleged theft (perhaps IBM has final control of the patents to whore out as they wish?).

And obviously it makes a difference to our speculation expedition whether we think Nintendo can just copy (for the most part) the MS tri-core design, or must use something fundamentally different. I rather suspect the former.
I don't recall anyone ever alleging MS stole anything. The story was IBM owned the IP so they sold it to a second customer. It's a good business move on IBM's part if the technology was not exclusive to Sony. If we're talking about the same thing then calling this "stealing" is way off base.
 
I would imagine that anything new would take cues from recent PowerPC developments. Xenon is pretty simple, simpler than even Broadway in some ways (OOE for ex).
Calling a CPU that likely has 3x the number of transistors of another one 'simpler' is a bit far-fetched, be that with a 'in some ways' disclaimer : ) Moreover that the OoE of the 750cl was not that advanced (reorder acted only on the top three ops from the instruction fetch queue, IIRC). It's not like it resembled 970's OoE in any shape of form.

It's interesting that they stress both easy development and easy porting from PC and 360. That leads me to believe that it will be either similar to PC or similar to 360. So maybe Llano is an option? I doubt they can use anything closely resembling 360 architecture without MS being unhappy.
Why would MS be unhappy about IBM technology going into the products of IBM's clients? Also, Llano's not an option in a console said to be backward compatible with the wii ; )
 
Calling a CPU that likely has 3x the number of transistors of another one 'simpler' is a bit far-fetched, be that with a 'in some ways' disclaimer : ) Moreover that the OoE of the 750cl was not that advanced (reorder acted only on the top three ops from the instruction fetch queue, IIRC). It's not like it resembled 970's OoE in any shape of form.
Well Broadway is one core and Xenon is three. ;) So I guess that means the cores are probably somewhat similar in transistor count then.

They went for 3 simple cores that brought in a lot of potential throughput but to keep it all cheap they limited the cache (6 threads, 1MB L2!) and kept some of the fancier CPU concepts away...
 
Well Broadway is one core and Xenon is three. ;) So I guess that means the cores are probably somewhat similar in transistor count then.
'CPU' was a slip of the tongue.. erm, fingers, I meant 'core'. PPC750cxe (750cl's closest sibling I have hard transistor data on) weighs at ~20M transistors. The PPE as found in Cell - ~60M. For the sake of ballpark figures we can assume a rough equivalence between PPE and Xenos' cores.

They went for 3 simple cores that brought in a lot of potential throughput but to keep it all cheap they limited the cache (6 threads, 1MB L2!) and kept some of the fancier CPU concepts away...
At the same time IBM thew in the most advanced SIMD they could for that money - VMX128 (I can't recall when VSX came to be, but I think it was a later development).
 
Didn't some IBM guy accidentally let slip that they had the next gen Sony and Nintendo contracts last year?
 
Last edited by a moderator:
Ok similar if not more powerful than 360/PS3 is a given.

"What we see is that we will be able to leverage a lot of the work we do for Xbox 360 and PS3 while we create games for the platform.

"So we will not have to redo completely the games that we create. We'll be able to use all the capacity the console is giving but also use all the work we do for the other platforms."

http://www.eurogamer.net/articles/2011-05-12-ubi-to-leverage-360-ps3-work-for-cafe

At least the usual (Ubisoft) multi platform games will be on Wii 2 too...
 
Last edited by a moderator:
Guys I am playing through Twilight Princess on both GameCube and Wii--I cannot help but to think what EAD could do with a huge leap in graphics processing power. Twilight Princess is still a wonder to behold artisically and in level design concidering how old our Nintendo console's graphics architecture is--Currently 12 years old. Just think what EAD could do in an HD Zelda with slightly/somewhat more power than Xbox 360 and PlayStation3.

^__^
 
"So we will not have to redo completely the games that we create. We'll be able to use all the capacity the console is giving but also use all the work we do for the other platforms."

This suggests to me:

512 RAM

Slightly more shaders (320).

Using all the work for other platforms=same assets=same RAM.

Use all the capacity of Wii2=bolt on slightly better shaders.

All falls in line with my predictions, 320 SP HD4670, 512 RAM.
 
Back
Top