Predict: The Next Generation Console Tech

Status
Not open for further replies.
Not sure how important BC will be, especially for Sony since they weren't the market-leading console of the previous generation.

What were the biggest games of this generation? Mostly shooters from 3rd-parties, if I'm not mistaken.

Would the ability to play COD-whatever make a difference in next-gen sales? Or would more people be interested in playing the latest COD developed for new hardware?

BC certainly wasn't a big draw in this gen, or at least not since the first couple of years.

What seems to be new somewhat is the multifunction nature of consoles. PS3 and Xbox 360 are both used by a lot of people as Netflix clients. In a world where smart phones and tablets are exploding in sales and potentially offering "good enough" capabilities in a lot of different feature-sets --- for photography, for media playback, for web browsing and yes, even for gaming -- consoles may be forced to try to boost non-gaming features.

Towards that end, maybe next-gen design should consider things like storage or connectivity to media servers, RAM for running multiple apps, perhaps some kind of SDK for developing non-gaming apps.

All of these things would of course eat up a bigger chunk of the BOM, leaving less to dedicate for gaming performance.

Designers may also consider trying to head off potential competition with mobile devices by enabling integration with phones and tablets. For instance, write iOS and Android apps. to control at least the non-gaming features of the consoles. Sony could have written an iPhone app. to control the BluRay playback rather than make people use the DS3. They could also look at using tablets for controlling certain types of games like RTS games.

They may not be able to afford to approach next-gen design the same way they have been doing for the past 2 decades.
 
When considering the importance of BC one should probably take into consideration PSN & XBLA content over retail...

I'm sure most consumers would be considerably bemused if they couldn't download and play their download titles immediately & instead had to re-purchase them all again...
 
Apologies for the pimpage but if anyone's interested I posted an article on #AltDevBlogADay on a similar subject...

I think you need to amend your post to include mentions of the upcoming PS4 tablet/console hybrid or Shifty Geezer would be very displeased and he runs this town like the Mafia. :p

One thing I would like to ask is how an upgraded Cell processor would handle some of the JIT compiled languages which are becoming more prevalent in development like Java? If the shoe was on the other foot and you were coming from the perspective of a small iOS/Android development house which was looking to make its' content ubiquitous with consoles, would you still have the same perspective?
 
Designers may also consider trying to head off potential competition with mobile devices by enabling integration with phones and tablets. For instance, write iOS and Android apps. to control at least the non-gaming features of the consoles. Sony could have written an iPhone app. to control the BluRay playback rather than make people use the DS3. They could also look at using tablets for controlling certain types of games like RTS games.

Yeah I'm hoping someone does this. Would be cool to use my iPhone as a fully programmable LCD remote control instead of a game controller.
 
Not sure how important BC will be, especially for Sony since they weren't the market-leading console of the previous generation.

What were the biggest games of this generation? Mostly shooters from 3rd-parties, if I'm not mistaken.

Would the ability to play COD-whatever make a difference in next-gen sales? Or would more people be interested in playing the latest COD developed for new hardware?

BC certainly wasn't a big draw in this gen, or at least not since the first couple of years.

What seems to be new somewhat is the multifunction nature of consoles. PS3 and Xbox 360 are both used by a lot of people as Netflix clients. In a world where smart phones and tablets are exploding in sales and potentially offering "good enough" capabilities in a lot of different feature-sets --- for photography, for media playback, for web browsing and yes, even for gaming -- consoles may be forced to try to boost non-gaming features.

Towards that end, maybe next-gen design should consider things like storage or connectivity to media servers, RAM for running multiple apps, perhaps some kind of SDK for developing non-gaming apps.

All of these things would of course eat up a bigger chunk of the BOM, leaving less to dedicate for gaming performance.

Designers may also consider trying to head off potential competition with mobile devices by enabling integration with phones and tablets. For instance, write iOS and Android apps. to control at least the non-gaming features of the consoles. Sony could have written an iPhone app. to control the BluRay playback rather than make people use the DS3. They could also look at using tablets for controlling certain types of games like RTS games.

They may not be able to afford to approach next-gen design the same way they have been doing for the past 2 decades.


I think it's a pretty forgone conclusion that the mulitmedia apps and abilities for the next gen of HD consoles (PS4 and XBOX 3) will increase.
Consoles are not just for gaming anymore and never will be again.
Not only is 3D looking like it's here to stay, but I'm expecting 1080p on top of that @120frames (3D=60 + 60).
But also think about things like PVR/DVR, for recording gameplay (or other self made movie clips) with audio and video editing and of course uploading to various sites.

The next set of consoles will have and need 8 gigs of ram (possibly even more) to do all of the things that are demanded....things that have been mostly the PC's domain.
the consoles can do some of those things now, but I expect these features and abilities to expand exponentially.



Oh and as archangelmorph touched on, backwards compatibility for the PS4 will not really be an option as it was with the PS2, because now the PSN has to be accounted for. Not only are there too many people invested in the network and all of the content they've purchased, but so is Sony.

PS4 will pretty much have to be compatible with PS3's content.
 
I think you need to amend your post to include mentions of the upcoming PS4 tablet/console hybrid or Shifty Geezer would be very displeased and he runs this town like the Mafia. :p

One thing I would like to ask is how an upgraded Cell processor would handle some of the JIT compiled languages which are becoming more prevalent in development like Java? If the shoe was on the other foot and you were coming from the perspective of a small iOS/Android development house which was looking to make its' content ubiquitous with consoles, would you still have the same perspective?

From such a perspective you're likely going to have limited content that performance-wise would hardly tax most modern consoles let alone next gen ones.

In which case I would imagine you'd be perfectly within your ability to completely ignore SPU programming & get your game up and running on just the 2 PPU threads & whatever GPU you have available.

Also I wouldn't expect a JIT compiled language like Java would run significantly worse on SPUs than on a more PC-like core (depends entirely on the code you're running though as well how optimal your virtual machine is as in some cases it could very well run faster).

However I wouldn't expect any developer in their write mind to attempt to write Uncharted in Java, running on a PS4 for example.
 
there is a java VM for SPUs, maybe someone wants to check out to estimate the performance
http://forums.ps2dev.org/viewtopic.php?t=11188

on a slightly related note, I was always wondering why Sony/IBM didn't write/create an emulator/VM on spu for ppu instructions. Using most of the SPU local store for JITted instruction cache and fibers to hide memory access, I think that could run quite well and it could reach up to 25% of performance of the ppu per spu (2threads vs 2threads), based on what I saw from other emulators.
Don't get me wrong, I love writing SPU assembler, but I think due to time constrains a lot of developers don't get the chance to get into it, then it would really give something to just get X360 CPU performance without much effort.
 
there is a java VM for SPUs, maybe someone wants to check out to estimate the performance
http://forums.ps2dev.org/viewtopic.php?t=11188

on a slightly related note, I was always wondering why Sony/IBM didn't write/create an emulator/VM on spu for ppu instructions. Using most of the SPU local store for JITted instruction cache and fibers to hide memory access, I think that could run quite well and it could reach up to 25% of performance of the ppu per spu (2threads vs 2threads), based on what I saw from other emulators.
Don't get me wrong, I love writing SPU assembler, but I think due to time constrains a lot of developers don't get the chance to get into it, then it would really give something to just get X360 CPU performance without much effort.

On a similar note,

Housemarque did a really interesting presentation on how they implemented their gameplay code to run in parallel on the SPUs...

Worth a look...
 
Only some questions to be put under discussion...so we can then say if the cell is well accepted by developers and if had a chance it evolved to a larger scale (more clock, 16 SPUs, 2-4 PPUs more efficient in single thread, with more cache, less latency, OOE +) and would be best option for sony (performance/wattage/size)?

Could it be Sony / IBM and AMD working together to provide an SoC Cell + GPU(or APU?)?

What would be more acceptable for developers cell or powerpc A2*?

* At 45nm only 428mm^2,1.43 billion transistors...maybe cell could be better relations performance/transistor/size/wattage with same transistors count(if not my mystake cell ps3 PPU + 7 SPUs 228/235 millions transistors has size something like 110/120 mm^2 at 45nm).

http://en.wikipedia.org/wiki/PowerPC_A2


Off topic: interesting file/article about Cell cluster /streaming processing
http://people.cs.vt.edu/butta/docs/ipdps09-cellmr.pdf
 
Last edited by a moderator:
A2 as it is is very server oriented. Cell would be more appropriate right now (aside from double precision).
 
A2 as it is is very server oriented. Cell would be more appropriate right now (aside from double precision).

I have never throughout my career in games so far ever seen a codebase that uses double precision for anything...

I can see it being useful for massive world coordinate systems a la GTA (even then there are alternatives) however I would never expect any dev studio to do anything beyond the absolute minimal function, called once per frame with it.

If anything it would be more favorable to see a chip that supports half-precision floating point natively (w/ 8 wide SIMD vector units in half mode), just like the GPUs...
 
A2 as it is is very server oriented. Cell would be more appropriate right now (aside from double precision).

Please correct me if I'm wrong ,FP16 Texture filter does not require 64bit / DP? The Cell with DP more effective would be useful to help gpu in this tasks(thinking something like APU)?
 
Last edited by a moderator:
The key is in the name... ;)
Thanx;) you are the guy!
16,16,16,16 -> RGBA

But cell with DP(sorry im still have cell im my mind... i dont think sony has spent billions in factories and development to throw in the trash ...) could be usefull for help in texture department like we see in frostbite 2 ?

We see many talks,benchmarks about nvidia have better texture filter FP16 than AMD chipset etc but thats another story...texture filter FP16 is the new paradign even for next gen console(more RAM,more bandwidth etc)?
 
Last edited by a moderator:
Status
Not open for further replies.
Back
Top