Predict: The Next Generation Console Tech

Status
Not open for further replies.
Looks like the next 15ish months are going to be awfully slow and boring then for the most part... Some excitement of course from discussion about new hardware and software capabilities, but most of the intervening time will be hum-drum monotony and a slew of ho-hum current console releases that will mostly fail to really impress technically (although not neccessarily artistically).

Hopefully Wuu will be able to fill that void to some extent!

What has been said about Wuu backwards compat? It'll run all Wii software, yes?
 
Looks like the next 15ish months are going to be awfully slow and boring then for the most part... Some excitement of course from discussion about new hardware and software capabilities, but most of the intervening time will be hum-drum monotony and a slew of ho-hum current console releases that will mostly fail to really impress technically (although not neccessarily artistically).

Hopefully Wuu will be able to fill that void to some extent!

What has been said about Wuu backwards compat? It'll run all Wii software, yes?

Wii U doesn't sound interesting, performance-wise.

But from a brief time playing around with Vita and playing some games on my iPad, I think there are opportunities to innovate on game mechanics for certain types of games with multitouch gestures.

Frankly, I'd be more interested in using multitouch than most motion gestures, because it can be more precise. I don't know how many touch points or what kind of screen the WiiU tablet controller will have but it'll be interesting to see if game designers are able to blend in a good mix of touch gestures with the physical controls.
 
Wii U doesn't sound interesting, performance-wise.

But from a brief time playing around with Vita and playing some games on my iPad, I think there are opportunities to innovate on game mechanics for certain types of games with multitouch gestures.

Frankly, I'd be more interested in using multitouch than most motion gestures, because it can be more precise. I don't know how many touch points or what kind of screen the WiiU tablet controller will have but it'll be interesting to see if game designers are able to blend in a good mix of touch gestures with the physical controls.

Wii U is single touch and non-capacitative, like DS and 3DS. Unless they're changing it last minute. It's received a fair bit of criticism for it. It may be pressure sensitive which could be interesting, but I haven't heard a definite word on that.
 
Wii U doesn't sound interesting, performance-wise.
Not in absolute terms, but if the hardware is very efficient, like Gamecube for example with its big cache and OoOE CPU with 1T SRAM, it could still create some impressive imagery.

Also, it'll be the newest console for a while, which automatically makes it more interesting IMO.
 
I think I already said it in another thread : "Photorealism is boring, but art is beautiful."
You want power to allow new gameplay, not to get failed photorealism.
 
I think I already said it in another thread : "Photorealism is boring, but art is beautiful."
You want power to allow new gameplay, not to get failed photorealism.

I think Nintendo games with PS3\Xbox 360 + level graphics are going to be beautiful.
 
I think I already said it in another thread : "Photorealism is boring, but art is beautiful."
You want power to allow new gameplay, not to get failed photorealism.

Photorealism has not "failed." It is also not boring.

The fact is photorealism is exciting. If it were not exciting, all major movies would just be rendered with a single high end gaming rig, and there would be no actors.

In my opinion, the first video game that looks truly photorealistic will sell millions of copies. Crysis, at the highest settings, comes somewhat close. I think that a high end console with a top notch GPU (for example a full GTX 680) and plenty of ram (minimum 4 gigabytes) will get even closer.

Right now, there is not a single video game that has the same level of realism as a TV show. When I can literally play through an episode of Star Trek Voyager (and it looks 99.9% as real as the TV version) I'll be very excited.

I think the next generation of consoles will be very powerful and will allow for games that are much closer to photorealism.
 
<snip>

In my opinion, the first video game that looks truly photorealistic will sell millions of copies. Crysis, at the highest settings, comes somewhat close. I think that a high end console with a top notch GPU (for example a full GTX 680) and plenty of ram (minimum 4 gigabytes) will get even closer.

Right now, there is not a single video game that has the same level of realism as a TV show. When I can literally play through an episode of Star Trek Voyager (and it looks 99.9% as real as the TV version) I'll be very excited.

<snip>

I'm not sure I agree with the bolded. But I would say that realism, not in terms of graphics, but rather in terms of animation, and simulation would be monumentally game-changing.

Being able to have proper fluid dynamics in games, fully simulated thermodynamics and chemical reaction kinetics, as well as advanced physics with fully destructable environments would make games completely different. The sheer level of interactivity you could have within a gaming sandbox would be epic.

I just don't know whether we'll even get to that point though, as I have no idea if HW can be designed to run these kinds of things in realtime alongside gamecode. My best guess would be not for the next 50 years
 
there are problems with this, you make a cave's entrance collapse so you can't enter it or exit it anymore ; vital plot devices get destroyed by accident (or not) regularly or otherwise inacessible, you blow a tire so you have to walk 30 minutes instead (back and forth). you decide to destroy a billiards table with an axe and NPCs go on talking to you as if nothing happened - but they need to check that the table has been destroyed so they don't keep trying to "play".

this can be a lot of fun but the combinatorial explosion is quite a burden (barring throwing you a MISSION FAILED! screen every time you do something wrong :LOL: )

ah yes, why not in the far future :yep2:
 
Right now, there is not a single video game that has the same level of realism as a TV show. When I can literally play through an episode of Star Trek Voyager (and it looks 99.9% as real as the TV version) I'll be very excited.

I don't know too much about star trek voyager, but I've been watching a lot of stargate recently (SG1 then atlantis), it's entertaining but playing it in a really realistic setting would reveal even more all the plot holes and incoherence maybe.

you walk in an unknown enemy base or ship (or the most important place of their entire planet, in a huge futuristic city) then steal an incredibly rare and powerful device the size of a shoe box which provides more energy and power than 1000x all the earth's nuclear reactors combined. but it's protected by two idle guards..

not counting the 150 times in the series were the bad guys would have succeeded taking over the entire galaxy if they had put a password on their control computers.
 
there are problems with this, you make a cave's entrance collapse so you can't enter it or exit it anymore ; vital plot devices get destroyed by accident (or not) regularly or otherwise inacessible, you blow a tire so you have to walk 30 minutes instead (back and forth). you decide to destroy a billiards table with an axe and NPCs go on talking to you as if nothing happened - but they need to check that the table has been destroyed so they don't keep trying to "play".

this can be a lot of fun but the combinatorial explosion is quite a burden (barring throwing you a MISSION FAILED! screen every time you do something wrong :LOL: )

ah yes, why not in the far future :yep2:

But then sure you can design around stuff like that in a way that makes sense like:
a) making the cave entrance out of metal, or a non-destructive material. This all depends on the tools you give to the player and how powerful you enable them to be. If you don't want the player to blow up the cave entrance, you can also not give him the dynamite with which to do it ;-)
b) Re the billards table, you can make a bouncer on the door remove the axe from the player so that he can't even take it into the pub to wreck things. But then when programming NPC AI routines, surely a simple check to see if billards table exists before running "play billiards" animation shouldn't be much of a problem, especially considering your have the platform power to do full destruction simulation in the first place.

I suppose you are of course correct, and the NPC AI would be a nightmare to program. But on the otherhand things like destruction were done in varying measures this gen, and even an evolutionary step beyond that would mean a step change in videogame interactivity and the scope for truly emergent gameplay. As things are currently, it seems things are going the other way and devs/pubs are throwing away physics and simulation for fancier graphics tech (see RF:Guerilla to RF:Armaggedon).
 
There's a thought actually... especially since this line of discussion is getting a bit OT and i'd like to lead us back on topic...

If Sony and MS really wanted devs to push things like simulation, animation (precedural and physics-based) and much smarter NPC AI, what kind of HW would be required for that?

... more in terms of data access patterns and whether such workloads would be better suited to GPU/GPGPU or a beefy CPU, thus informing the choice of the balance of silicon and transister budgets dedicated to each chip.
 
in Kingpin no arms were allowed in the pub :)

for AI the consensus seems to be you need strong and robust single-thread performance (lots of branching and conditions in the code)

for physics, you can have a lot of computations running on the GPU, but it may typically be cloth or particle physics that make for a good show but don't affect game physics. you could say this mainly uses "local data" whereas AI and game mechanics need to deal with global data.

AMD's "FSA" allows CPU and GPU to work more easily with each other's data. naïvely I can imagine a scenario where GPU simulations results are uploaded to the CPU so that game logic can acknowledge, and make decisions.
(Intel and nvidia, and others actors will be heading towards somehting similar)

of course the software part will be incredibly hard, and the hardware as well - eventually the data movement is the expensive part rather than computation. the internal and external (memory) bandwithes and latencies become the limiting and power hungry aspects (just like you can rate low end graphics card by their memory bandwith and almost ignore everything else)

this stuff will work out better on the generation after next-generation, but next-gen may pioneer it
 
DaE posted this. I'm not sure whether to laugh or not.

https://twitter.com/superDaE/status/233754001693343746/photo/1

Az52BqxCAAAEPLN.jpg


Supposed to be Kinect 2 on Durango.

Bigger pic link (can make out a lot more info)

https://p.twimg.com/Az52BqxCAAAEPLN.jpg:large
 
have you posted this on gaf yet?

nope. you want too?

trying to figure out the best way to elicit more info. he seems to want more followers, so if i posted a new thread on gaf it might lead to a flurry of new followers for him. assuming thats a good thing or not. otoh maybe attention would scare him to silence.

i could dump it in that recent next gen what we know thread if i can find it. i'm counting on that being the next gen scuttlebutt thread for gaf going forward. It probably wont get much attention there though.

or, i could always do nothing.

If that is really Kinect two, well, I guess we can clearly make out fingers now.
 
nope. you want too?

trying to figure out the best way to elicit more info. he seems to want more followers, so if i posted a new thread on gaf it might lead to a flurry of new followers for him. assuming thats a good thing or not. otoh maybe attention would scare him to silence.

i could dump it in that recent next gen what we know thread if i can find it. i'm counting on that being the next gen scuttlebutt thread for gaf going forward.

or, i could always do nothing.

I'll make a thread.
 
nope. you want too?

trying to figure out the best way to elicit more info. he seems to want more followers, so if i posted a new thread on gaf it might lead to a flurry of new followers for him. assuming thats a good thing or not. otoh maybe attention would scare him to silence.

i could dump it in that recent next gen what we know thread if i can find it. i'm counting on that being the next gen scuttlebutt thread for gaf going forward. It probably wont get much attention there though.

or, i could always do nothing.

If that is really Kinect two, well, I guess we can clearly make out fingers now.

More than that, it seems it is able to see fingers even when your hand is in contact with your body... Considering how thin a finger can be that can say a lot about the precision of the new camera.
 
Status
Not open for further replies.
Back
Top