Sony's console platform policy

Status
Not open for further replies.
In the light of PSP spec revelation, is anyone else not disturbed by SCEI's lack of a consistent platform policy????

I mean, SCEI had the option of taking either PSX1 or PSX2 architecture, tweak them for mobile gaming, and release it as PSP. This would have spared developers from a lot of trouble and expense of having to learn a completely new platform.

Instead, SCEI chooeses to reinvent the wheel, developing an entirely new micro-architecture that seemingly offers little benefit over an existing solution, namely the PSX2 microarchitecture, why must SCEI do this every time?? Soon, Kutaragi will unveil the PSX3 microarchitecture, which itself is entirely different from PSX2 and PSP.

Why must Kutaragi reinvent the wheel every time when existing wheels could have been just as effective with slight alterations????
 
because technology moves quickly? oh and there is no fallback (DX nor OpenGL) to rely on. I agree that this would soften transistion periods for devs but just how feasable is this?

I guess we will see with regards to XBnext how beneficial maintaining a set of tools across generations is.
 
...

Kutaragi's vision includes a CELL powered handheld. Does this mean SCEI is going to dump PSP in 4~5 years down the road and come up with yet another completely incompatible portable????

With MS, there IS a platform policy and developers can expect the pretty much same environment in Xbox2, with bits added here and there. Why can't SCEI be the same??? What is wrong with "standard" anyway???
 
Simply put because Sony knows its business then anyone else and it has determined that a new architecture is worth it. Whatever their reasoning was I am sure it was thought out by them for a long time. They could have shrunk the PS2 architecture down to make it portable, but do they really need development of portable games to go through the roof? Also, it might be going out on a limb here but maybe this new architecture wouldn't run so hot and would provide much longer battery life.

I wouldn't consider it reinventing the wheel either. Just developing a new architecture that is forward thinking and could be used for years upon years as a handheld device that only gets cheaper as the years roll by. Maybe after 2 years of production the thing will be vastly cheaper than a portable PS2 solution would have been.

All factors should be considered before claiming that Sony lacks a console plan.
 
trial and error designs

nah that would mean that they keeps the designs that work.


I mean, SCEI had the option of taking either PSX1 or PSX2 architecture, tweak them for mobile gaming, and release it as PSP. This would have spared developers from a lot of trouble and expense of having to learn a completely new platform.

fair enough for the market bracket we are talking about. wouldn't it be just as effective to migrate dev tools across platforms?

Why must Kutaragi reinvent the wheel every time when existing wheels could have been just as effective with slight alterations????

cause he's a looney, I though we'd established this?[/quote]
 
Deadmeat your comparision is not quite as accurate as you might think. You mentioned that Sony had the option of shrinking down PSOne or PS2 and putting it on a portable. Why? The PS2 was clearly a rushed architecture and PSOne is ancient, and neither supports any technology that was introduced after 1997. Secondly, neither chipset was built for running on a very tight power budget. PSOne, to this day, is still rated at about 8 or 9 watts even after AC-DC conversion, enough to kill 6 AA batteries in three hours (with no screen). It's like wanting to build a fuel efficient car by just "shrinking" a '68 Corvette. You seem as if you have an EE background, you should know this. :?

Portable architecture has often been quite different from console. Game Boy wasn't a Nintendo, Neo-Geo Pocket certainly wasn't an MVS, Atari Lynx wasn't any of Atari 16 bit computers. The only times it has been similiar is where the portable is either the exact same system as the home version (PC-Engine Express->PCE) or virtually the same (Sega Game Gear->SMS).

And as for the "dumping" of PSP in 4-5 years we can only hope that as lithography improves and new 3D algorithms are made practical for use on ASICs that we will recieve better and better systems. Besides, it's not like Nintendo would ever "dump" a system *cough* GBC *cough*. :p [/b]
 
PlayStation 2 was Sony's first fully custom built console and the first one they would have had to manufacture by themselves ( one fab was shared with Toshiba, the first Oita plant ).

SCE's first task seem to have been the development of the GS which followed a GPU that not all developers were happy with: PSOne seems to be quite GPU limited ( fill-rate wise and triangle set-up wise ) and they felt the need for something MUCH better...
The feature-set for the GS was locked down ~1997 and its design was, in specs, earlier than the EE was...

The physical chips was not ready by that time as they probably did not have the technology to manufacture such a large chip with 4 MB of e-DRAM as of yet.

In all available EE docs there is some mention that the EE was designed to keep up with the enormous fill-rate they were planning to achieve with the GPU ( very clear in "Designing and Programming the Emotion Engine" )...

The problem was that while EE's development seems to have been smoother, the GS hit some road blocks in its way to production and their manufacturing process ( 180 nm ) initially had lots of troubles fitting such a large chip as the GS hence the PlayStation 2 consoles that shipped with hot and large 250 nm GS chips.

There is a reason why Sony and Toshiba have been working VERY hard on next-generation manufacturing processes, learnign as many tricks from IBM as they could and the result is a 65 nm process earlier than most competitors and a good deal of work already done on their next step, 45 nm...

SCE has seen how much better PlayStation 2 production became when the manufacturing problems with the GS ended and when they put more resources on pushing the manufacturing technologies harder and harder while developing new ones... hence all the die shrinks they have been doing of both EE and GS and now we can enoy a 8 Watts 86 mm^2 EE+GS chip using 90 nm technology.

The EE is clearly the most modern component out of the two, the Vector Unit are still a design which is liked by most programmers for their flexibility ( > DX9 ) and RAW speed...

When they locked down the feature-set of the GS tghe programmable Shaders revolution had not started yet and Sony did miss it ( luckily for them the Hardware is still pretty powerful... hence we can admire Siletn hill 3 gorgeous yet creepy visuals :) )...

Sony has been much more active in collaborating with unviersities across the globes and paying attention to the development of 3D graphics, so they have learned few more things...

I went a bit off-topic, sorry...

IMHO it would have costed a lot to optimize the PlayStation 2 solution to very low power consumption and it might have crippled performance in indesirable ways...

A souped up PSOne would have not been worthy either... the only thing that would have not been modified too much would have been the MIPS core... and infact we do have nice R4000 chips in the PSP...

Why re-inventing the wheel ?

Deadmeat, you must not code or invent much... you would know that engineers/programmers always feel that they could create a better wheel ;)

And judging by the PSP specs, I think the decision the made was the right one.
 
Well they didn't really throw the PS2 arch out, the PSP does have its own VU. Whether or not it shares much with the PS2 VUs remains to be seen.
 
I'm just happy that they took seriously those early complaints about difficulty of PS2 programming and that they've making sure it won't repeat with PSP. It's only reasonable to expect they will approach the PS3 development issues with the same seriousness.
 
DeadmeatGA said:
In the light of PSP spec revelation, is anyone else not disturbed by SCEI's lack of a consistent platform policy????

Quite frankly, the only thing that disturbs me is seeing how much mindless fud, bs and negative anti-sony propaganda you spout on this forum.

Look, pal. Sony reinvents their platforms most likely because that is the best option for them. You see, a handheld designed for release towards end of 2004 does not have the same needs as a stationary console that was released early 2001. Technology evolves, and evolves very quickly in the semiconductor business.

Having sold more stationary consoles than ANYONE in this generation (and the previous one as well), I think they know just slightly better than you what their best options are.

Besides, I can SWEAR that you'd slam them for lazyness or whatever if PSP had turned out to be a re-worked EE+GS combo! With you, it doesn't matter what sony does, you find something to bitch about anyway (most likely because you're nothing but a friggin troll).


*G*
 
Deadmeat said:
In the light of PSP spec revelation, is anyone else not disturbed by SCEI's lack of a consistent platform policy???? I mean, SCEI had the option of taking either PSX1 or PSX2 architecture
So this isn't PSX2 like enough for you?
It's hybrid UMA, DMA controlled system, MIPS CPU with Vector units, fast embeded ram GPU, and dedicated hardware for movie playback.
It has same width system bus and same clock divider for Bus/CPU frequency. Going by these slides it practically screams "portable PS2 with improvements".
Eg. - even from this incomplete spec we can see they've fixed some of the more glaring inconsistencies/problems with PS2 graphic subsystem. Particularly inclusion of more of the "primitive" GPU functions like clipping, TC, etc.
 
Deadmeat I presume is effectively saying why Sony see the need to dump dev tools/middleware/etc... across each genetration. summed up why doesn't the console market a DX type paradigm?


or at least I hope that's what he saying.
 
chaphack said:
I think he meant, Sony should have CELLuslise the PSP. It be better for their mega CELLplans.

nope he's clearly pissed that Sony have no coherent (read legacy/modular) development tools across genetrations. and by implication Japanese hardware design seems to be his pet peeve.

what he avocates looks alot like the PC GFX paradigm with DX as an dev enviroment.

in short maybe consoles could benefit being more like PC's albeit with a 4-5 yrs upgrade cycle.
 
Deadmeat I presume is effectively saying why Sony see the need to dump dev tools/middleware/etc... across each genetration. summed up why doesn't the console market a DX type paradigm?
It's a fair argument but, in order to keep something you need to establish it first and they completely failed to do anything of the sort on PS2. (it could be argued they never even tried).

Although if that was his point he shouldn't so specifically cry why hardware design isn't minituarized PS2 , even though it is :p
 
It's a fair argument but, in order to keep something you need to establish it first and they completely failed to do anything of the sort on PS2. (it could be argued they never even tried).

thats right Sony doesn't try in this respect. if you look at the last few generation of consoles none even considered such an option.
 
Still there is some point that Sony should strive for a standard, not just a global standard, but a standard within themselves. UMD/DVD/BR/Vaios/Clies/PSX/PSP/Memcard/Memstick/etc. Many technological differences imho. Maybe CELL will do something about it. It be better for their wares and their BB ideals.
 
Status
Not open for further replies.
Back
Top