Follow Slashdot stories on Twitter


Forgot your password?

IBM To Make CPU For Sony's PS3 110

SmasKenS writes "So, not only did they get to make the CPU for Nintendo's GameCube, now they work for Sony too. Saw this on voodooextreme first, they also got a link to BBC News." Now, granted, this is several years away from actually happening, (projected date of 2004) but the costs (and profits) that are involved are staggering.
This discussion has been archived. No new comments can be posted.

IBM To Make CPU For Sony's PS3

Comments Filter:
  • by glenkim ( 412499 ) on Wednesday June 13, 2001 @09:11AM (#154643) Homepage
    Now we're gonna get "Peace, Love, and PS3" on our sidewalks.
  • what IBM will come out with by 2004. The amount of potential that is there for game consoles is staggering and I think IBM is one of the companies that can do some real damage here.

    I just hope you don't have to get IBM to service your PS3, that could take years.

    Murphy's Law of Copiers

  • by number one duck ( 319827 ) on Wednesday June 13, 2001 @09:14AM (#154645) Journal
    Well, I guess thats one way to crush the gamecube/xbox. With the marketting involved, just the very mention of 'ps3' is enough to make people shiver.
    Why buy xbox vapor, when you can buy ps3 vapor?
  • The platform isn't that complicated. The graphics might be, but you could incrementally improve those. And a place like Sony could crank out the manufacturing engineering in a couple of months.

    Must just want to spread out the innovation space to give the PS2 time to make cash.

    In three years, PS2 will be a board-game spinner compared to the cards available for desktops.

    Maybe they're looking for the next great parts shortage so they can order a zillion of those...

  • Slashdot readers love to preach about the evils of monopolies - so why does it feel so good to root for IBM getting its foot in the gaming industry door? What is it about IBM that we all secretly like? I can't believe it's the Linux push, because that's a recent thing. Maybe it's their slick notebooks.
  • as good as this might be for the PS3 platform, it's bad news for the people buying it. The more that these companies become intertwined with each other (i.e. Nintendo and Sony pulling from the same manufacturer), the less competition in the market. And less competition means higher prices. Higher prices mean this could potentially suck.

  • by number one duck ( 319827 ) on Wednesday June 13, 2001 @09:19AM (#154649) Journal
    I still don't understand what the hell 'virtual music' could be, but the commercials for the ps9 made it sound incredible. Now where's my soma...

  • I had heard that the PS3 was going to employ parallel-processing stuff, which was complicated (at least from the game developers point of view), and the GScubes were being sent out to let people get their head around how it was going to work...

    But, why would they rush it to market when there's plenty of good money to be made from the PS2 platform for the next couple of years? They need to get their money back from that platform first.
  • > The platform isn't that complicated.

    Spoken in true slashdot manner - you've obviously never coded on one.

  • IBM does, using a form of the PowerPC processor they call 'Gecko'(see here []). It's the x-box that uses Intel chips

  • I misread the post, I'm a idiot who can't read, sorry!

    -henry, the idiot :(
  • by OnyxRaven ( 9906 ) on Wednesday June 13, 2001 @09:25AM (#154654) Homepage
    I read through the posts that showed up here quickly and the posts over at voodooextreme and I think people are a bit confused.

    The fact that IBM and sony announced their partnership here doesnt mean that they're switching gears already for the ps3. All it means is that they are announcing the beginning of development for the platform. How long did the ps2 take to develop? the x-box? a few years. This is nothing new in the way of business relations.

    What I am excited to hear about though is the kinds of technologies that IBM is planning on using or hoping to use in their new chips... They have a few years to develop it so I assume they have some high goals!

    I also wonder what platform squaresoft is going to contract to next. will they stick with sony? renew their relationship with nintendo? join the evil empire (scary thought)?

  • The three companies aim to design a "super-computer on a chip" with a wide variety of consumer applications, they said in a joint statement. "The result will be consumer devices that are more powerful than IBM's Deep Blue super-computer, operate at low power and access the broadband internet at ultra-high speeds," the statement added.

    If IBM can really implement a supercomputer in the PS3, how many will Saddam Hussein buy?
  • Ok, here's what I know so far...

    The GSCube, Sony's high-end graphics workstation is supposed to be able to handle 64 sets of Emotion Engines/Graphics Synths.

    The PS3 is (or was) expected to be equivalent to 16 sets of EEs/GSs and is due sometime in 2005

    How does Intel's new microtransistor technolgy fit in? Will this unanticipated advance be duplicatable by IBM by then? I know that production chips aren't due until, when, 2007?

    Will the console maker that follows PS3 be able to sign with Intel for chip fab?

    I *MUST* know - The future of Gran Turismo depends on it!

    - Wi11 r4c3 ph0r b33r

  • So, IBM provide the chip for Nintendo and IBM provide the chips for Sony.
    I want IBM to provide a system that plays Nintendo & Sony games, by licensing the technology from both parties. Panasonic are providing a licensed GameCube(which looks like a Radio Alarm clock), so I can but hope.
  • by Rei ( 128717 ) on Wednesday June 13, 2001 @09:26AM (#154658) Homepage
    Seriously, though, it isn't the hardware that makes a console. What matters is who the companies that make games develop for. You could have the best system in the world, but if you don't get the Squaresofts of the world writing games for you, it won't amount to anything. While I pine for the days when writing your own 3d engine that beat all of the others was the mark of excelence, nowadays it is the artfulness and resourcefulness of the game design that matters. The key, now, will be attracting the developers to one side or the other. Part of that will be done by market forces, but part of it will also be done by how friendly the company is to developers. During the PS1/N64 competition, Sony showed they were superior in this aspect, as many of Nintendo's long-time supporters switched over to sony, even though their platforms were comparable (sony's had a little better overall performance for the cost, but not too much...). Anyways, we'll see what happens now, when the mists clear.

    - Rei
  • by nanojath ( 265940 ) on Wednesday June 13, 2001 @09:27AM (#154659) Homepage Journal
    From the article...

    " "The result will be consumer devices that are more powerful than IBM's Deep Blue super-computer, operate at low power and access the broadband internet at ultra-high speeds," the statement added."

    It's really great to see that all of IBM's investments in basic research in naoscale technologies will find their apotheosis in making Lara Crofts boobs jiggle EXACTLY like Angelina Jolie's. God I can't wait for the future to be here.

  • Very true that many readers fail to hold all market dominators in the same light. In this situation, however, IBM hasn't dominated the game market. Another note is that Microsoft abuses monopoly power and quite often sticks their head out of the window to tell the world who they just crushed or who they want to crush. (AOL situation? Netscape situation? GPL situation?)

    IBM was also investigated for monopoly power, and ironically, David Boise defended them successfully. Remember Mr. Boise as the prosecuting attorney that successfully prosecuted Microsoft on behalf of the Department of Justice. This ultimately falls under the age-old saying "the squeaky wheel gets the grease". Cisco, IBM, and AOL simply don't squeak like Microsoft.
  • Curious, consoles don't usually go in tiny increments as it takes a few years to convince companies to write good games for it. Good being high quality. The reason it takes them so long is that learning the ins and outs of a new platform is tricky. That said, considering all they have is a high level idea of what they want to develop 4 years could be considered quite short. When was N64 released? What was the gap from PS1 to PS2? How long did it take to get Mozilla from the start to finish? Why was Win95 three years late (Sorry, no proof of this -- just speculation). Sonys been good at releasing things on time (within 6 months can almost be considered on time) even if they do create artificial shortages to make it seem popular. I'm sure they know what they're getting into. BTW. The Itanium took how long to design and release? P4? How about the Athlon? Big projects are generally long term projects. Its people who bitch that cause poor results for the companies willing to sacrifice the long term cash haul on a good product for a short lived "we're going bankrupt and your unsupported because the owners ran with the money" type products.
  • <TROLL>I guess BBC News reads Slashdot too, then... []</TROLL>
    (I'm sorry, I just wanted to see what it felt like to b*tch about this fine site repeating itself)
  • I heard that the GSCube (which is PS3) would be built out of 16 PlayStation chips and helps developers do parallel processing using these 16 chips. If these were just PlayStation chips, then whats IBM doing different. If on the other hand, if this is a totally new architecture with IBM holding the specs to the new chip, which according to the article is faster than a Deep Blue, would be interesting to see.
  • Not to mention designed a processor -- let alone a processor using new techniques. Or a high speed bus (damn transmission line effects on those high frequency things), not anything else. I'd be willing to say this person has not a clue. I'm no wiz, but atleast I've done a good chunk of electronics work and programming to know just testing a new processor design takes 6 months to a year to work the bugs out and thats without any significant overhauls.
  • Well, i guess, having finished the PS2, Sony could just sit around doing absolutely nothing for a few years!
  • by Dutchie ( 450420 ) on Wednesday June 13, 2001 @09:33AM (#154666) Homepage Journal
    By God, that is just unbelievable. Consumer devices with that much power, in a GAME console for crying out loud. Ridiculous... but VERY COOL. I've said in some previous post already that I like IBM's style. Ofcourse I got told then that I am SO easily manipulated, but I'm still of that opinion. I mean, everybody knows that IBM's marketing has ultimately sucked in the past, otherwise Linux would now be competing with OS2. But I have to admit that having a rather poor marketing machine but a KICKASS research and development makes IBM still one of my favorite companies (ok, I kind of like Philips too, but their marketing doesn't just suck, it's downright horrible...). I hope superior technology will eventually beat the crap out of marketing waterheads that only know how to sell stuff and nothing about how to actually MAKE good stuff. Call me naive again.
    • Imagination is more important than knowledge.
  • by Anonymous Coward
    PS3? We'll never see it except on ebay selling for more than cars.
  • Or does even the promise of the spankiest Gfx on the planet, the fastest processor in the history of mankind, and the slickest flashing lights on the box have only about a tenth the impact of waiting for that new Megadrive (genesis) to arrive from Japan.
  • What, did you think that they'd sit there and go "Well, this should be the last platform we should ever have to develop..."?!

    It's called good business - make the best product you can with the best technology available today and sell it. While you're selling it, rely on the ever increasing level of technology advancement and start developing the next iteration.

  • by rtaylor ( 70602 ) on Wednesday June 13, 2001 @09:39AM (#154670) Homepage
    IBM's not exactly a monopoly -- atleast not since the trail. They're rarely #1 at something but they're almost always second or a close third in nearly everything to do with computers. DB2, Lotus, their hardware (harddrives, processors, etc.) are all good and sometimes they'll lead for a short while. I guess their trick is to develop the technology then license it off to competing companies. Without IBMs tech some companies couldn't keep up (wheres Quantom now?). Hence I don't consider them a monopoly in any field -- but they're a strong player in nearly all of them.
  • Intel spends billions on each new architecture.
  • I also wonder what platform squaresoft is going to contract to next. will they stick with sony? renew their relationship with nintendo? join the evil empire (scary thought)?

    As someone who buys consoles on the rule of thumb, "Which ever Square is supporting", I of course bought a PS2 in preperation of FFX. FFXI is also in development as a online only game, made specifically for the PS2. Beyond that, Square has indicated that they're planning on sticking with Sony for future projects and there really isn't any reason for them not to. The reason they bailed on Nintendo was due to Nintendo's reluctance to put a CDROM in their system, so long as Sony provides a viable platform Square has no reason to leave.

    Now, if a competitive game system gets enough users, Square may decide it's worth while to hire another team of programmers and try their hand at cross platform development. Really the XBox has to prove itself several million times over before they can establish the clout a console maker needs.

  • this is basically old news. [] the partnership was already there; the only new news is that this chip is going in the PS3. but c'mon, who couldn't predict that from the earlier story?

    as far as i know, sony's plan for the PS3 puts it more squarely in the net appliance/set top box market. this news is in line with that plan.

  • if there is a major breakthrough in quantum computing this year.

    That's the joy of the computer industry:
    a few years in the future, you could be using completely different technology.
  • at Fry's Electronics here in Houston. However, that part of town was hard hit by last week's flood so they might not be in sellable condition.
  • Speaking of cubes, and GameCube licencing of said cube tech, I can think of another company that makes a cube-shaped machine with a PowerPC-based processor, ATI graphics, and DVD-based storage capabilities. Apple licencing the GameCube tech to stick inside their machines would certainly a good way to stop people complaining about the lack of games on the Mac platform. The extra hardware cost would be minimal, as well...
  • That's a good point. There really isn't that much time between the releases to allow the console to mature. PSX had been around for a few years even before PS2 was mentioned in a public domain. Now PS2's been out for less than a year, and people are talking about PS3 in a big way. This is going to make console gaming like PC gaming in that there's a perpetual sense of planned obsolescence. I thought the appeal of console gaming was the stable hardware platform.

    However, this could be a sign of changing times, an evolution of console gaming, if you will. In that sense, it's interesting to see how console gaming will change from here.

  • by UnknownSoldier ( 67820 ) on Wednesday June 13, 2001 @09:50AM (#154679)
    Instead of IBM developing 2 chips, why couldn't they develop just one and send it to both Nintendo and Sony. Think they would notice? ;-)

    It *would* lower the hardware production costs if Nintendo and Sony ever "standardized" on commodity hardware. Nahh, they'll never go for it, it would make too much sen$e. ;-)

  • But extremely expensive :)
  • I want IBM to provide a system ...

    If you have to ask how much it would cost, you couldn't afford it.

  • Wasn't deep blue just a gaming console anyways? I believe all it did was play chess. ;-)
  • by s20451 ( 410424 ) on Wednesday June 13, 2001 @09:59AM (#154683) Journal

    Although IBM is a big bad company, they have retained "cool" status because of the following (in no particular order):

    • They embraced Linux.
    • They built a huge multiprocessor computer that kicked Garry Kasparov's ass.
    • They developed the IBM PC, which was expressly designed to be cloned, hacked, modified ... (unlike the Mac, which was always intended to be sterile and closed)
    • They actually innovate - IBM Research [] has come up with some remarkable advances in practical computing, as well as mathematics and other abstract fields. IBM labs sponsor pure research, unlike many other large companies.
    • They have already been chastened with antitrust suits, and appear to have learned their lesson.
    • They're the grand-daddy of computing.
  • Check out con_tech.shtml for some detail.
  • Heh I quite agree with the square-support rule as well. FFX is great - played it (in japanese...) at E3 and was thouroughly impressed already, even in their limited demo. XI looks interesting as well, even though not a whole lot is known about it yet.

    As for the Xbox. I was actually just discussing this with my friend (Xbox zealot) and even he is questioning some of the decisions and strange information leaks coming out of the MS-Xbox camp. Such things as Xboxes being shipped without any broadband hardware, dvd support (ie you have to buy the remote!), and a slew of other wierd problems.

    The broadband issue has already come up. The only game that has kept my interest in the xbox, Halo, is NOT being shipped with support for broadband. They were not forthcoming on what exactly this means, but I assume it's that they didnt get the networking SDK in time for the game to be released on shipping, but the hardware is still in the xbox for when they update the game. (is everyone ready for patches for their console games? Yay, great.) Scary thought being that they will be releasing a broadband addon (100$ probably) later to enable the support (so 400$ for a full networked gaming system, whereas both sony and nintendo are shooting for 300$ max).

    This all comes down to microsoft's release-wait-see-patch-wait-see-fix-repeat scheme of everything - "lets release the xbox as soon as possible, and wait to see how well it does, then we'll add stuff/fix problems. yah." I dont think that is going to fly in the loyalty-driven console market. People are going to realize that microsoft is really just along for the ride, and the xbox will go the way of the Jaguar (good tech, bad implementation).

    But all of the above is obviously pending some offical release of info, or maybe even the actual hardware.

    BTW the xbox hardware at E3 was relatively interesting... midtower boxes with dvd drives and nifty faceplates. It was said a couple times that it was only really running at half effeciency - mainly that the sound was being processed in software for some reason.

    ... ANYWAY ... i guess this ends my way-offtopic rant ...
  • Wasn't deep blue just a gaming console anyways? I believe all it did was play chess. ;-)

    Ask Kasparov, I think he'll call it the daemon machine instead :P

    • Imagination is more important than knowledge.
  • they do already do this. its called the powerpc. the same g4 in macs is in the gamecube
  • If this one is going to be made by IBM it will propably be some kind of PowerPC in a future version, and this means that OSs like Linux, NetBSD and OpenBSD will be able to run on it with little modifications, and that just sounds like something that i could go for.
  • Angelina's boobs don't jiggle that much - they're in a padded bra... lara croft can continue to be better than the real thing... insert bitch about sexism and objectification of women here-
  • I didn't ask how much. #8^P
    Anyway, you've just reminded me of Irn-Bru, which i can only get smuggled into this country. Hmmm, Irn-Bru, Salt & Vinegar crisps & Cola Bottles. Drool...
  • It seems that Sony is starting to do what sega did, which is making a new system every year or two. Now we all know that Sega didn't do to well with that plan, but if Sony can keep the newest PS compatible with older PS games, they might do better than Sega.

    And this comment comes from a person who doesn't own any Sega or Sony game consoles.

  • The new chip will be designed for the broadband era, allowing the games machines and other "intelligent" devices to communicate with each other or connect to the internet.

    The three companies aim to design a "super-computer on a chip" with a wide variety of consumer applications, they said in a joint statement.

    The chip will also be capable of massive parallel processing - dividing up complex or time-consuming processing tasks among many chips - and could eventually be used in computer products.

    Oh, yeah, it'll play video games, too.

  • Sony To Do list 1. Games to Graphically Soup up for PS3!!! Burger Time Ikari Warriors Pole Position Qbert
  • Console manufacturers sell their systems at a loss. They only make profit from game royalties. So it's to their advantage that games for a console are produced for as long as possible. And since console hardware architectures change so much, it takes a long time for game makers to get the hang of new systems. And since the games don't work on older systems, lots of console owners get pissed off that they don't work on systems they bought just a year ago.

    In the PC world, new processors come out all the time. But new software will still work on 3 year old systems (typically). There are plenty of API and driver layers to make hardware transparent. But on consoles, there isn't much of an OS. You work on bare hardware basically. Games then become very dependant on the behavior and timings of the hardware. The advantage is they can squeeze the most capability out the hardware and use very small amounts of memory. The disadvantage, of course, is that forward compatibility is not possible without emulation.
  • by Chris Pimlott ( 16212 ) on Wednesday June 13, 2001 @11:01AM (#154695)
    This isn't a new role for IBM at all. IBM was contracted to make and assemble Atari Jaguars. As most /.er know, they fab Transmeta's chips too. IBM one of the best fabribation capacities in the industry, so it should be no surprise they would built PS3's.
  • The new chip will be designed for the broadband era, allowing the games machines and other "intelligent" devices to communicate with each other or connect to the internet.

    Sounds like the Intel ads, "better Internet experience". How a different chip allows the above is complete crap, its the software written for the chip that does that!

  • Yeah... too much sense and not enough cents ;)
  • wrong. Its not a G4 in the gamecube. The g4 was originally motorola (ya know...velocity engine and all that stuff). IBM wanted to concentrate more on getting the Mhz up on the powerpc chips and motorola wanted to throw more features in. That split the group apart. Now IBM is taking on some of the fab for motorola since their engineers are incompentant. The powerpc chip in the gamecube is an extension on the powerpc family, but it is NOT a G4.
  • Great so only Garry Kasparov can beat any of the games...
  • by _xeno_ ( 155264 ) on Wednesday June 13, 2001 @11:17AM (#154700) Homepage Journal
    I'll accept most of your points except:

    They developed the IBM PC, which was expressly designed to be cloned, hacked, modified

    Bullshit! It was not meant to be cloned, hacked, or modified! Remember how Compaq got their start? By making a clean-room copy of the IBM BIOS. In the early days of the PC, IBM was just as tight about their hardware as Apple was. Eventually, though, other companies started copying them and the PC was born. That "commodity hardware" didn't really start until the mid to late 80s, after Compaq had successfully created a PC clone. (I think - my dates may be wrong.)

    Eventually IBM had to let go of the PC, but it wasn't until after they were stung with anti-trust lawsuits.


  • In a seemingly unrelated story, Saddam Hussien walked into a Electronics Boutique store in Baghdad today and preordered 1,000 Sony PS3's. Chris "Tesing proves, testing works!"
  • early as 3 months after the launch of the psx and it still took 5 years.

  • You forget the 1 feature that will appear on the PS3 for developers and gamers alike: a small expansion bay right on that side to hide your weed in.
  • ....until at least 2005, I'm really not that worried.

  • "People are going to realize that microsoft is really just along for the ride, and the xbox will go the way of the Jaguar (good tech, bad implementation)."

    The Jaguar was not good technology. The statement of it being 64bit was a lie. It consisted of multiple processors that added up to 64bits total. That is why the games did not look impressive. That piece of shit was barely more powerful than the ill-fated 32X and was nowhere in the league of the Saturn, PSX, and N64.
  • Didn't Sony make a bunch of fabs for themselves a while ago to put out vast quantities of EE's and GS's? What's going to happen to them? It seems like it'd be an AWFUL waste of money to create a fab for just one generation of chips...
  • Maybe Apple should use these for the newer Macs. Anything is better than trying to get those chips in short supply/high cost from Motorola....
  • you are almost correct. Try, not even Gary Kasparov can beat the games. ;-)
  • Did anybody read the article ... Chip Production doesn't start until 2004. How can this be marketing vapour? Who the hell will wait over 3 years for the best console.

    The previous post wasn't great, but shit , it was on topic and wasn't false.
  • Cingular Wireless has just started a billboard campaign the irritates the hell out of me. The billboards have a peace sign, a smiley face, and the Cingular logo. Underneath it reads: "Peace, Happiness, Cingular."

    Gee, how creative...
  • IBM will re-introduce the Microchannel bus, claim that the PS3 is the next generation IBM PS2...

  • This isn't a new role for IBM at all. IBM was contracted to make and assemble Atari Jaguars. As most /.er know, they fab Transmeta's chips too. IBM one of the best fabribation capacities in the industry, so it should be no surprise they would built PS3's.

    I don't know why the parent's being moderated as funny. The Jaguar portion is 100% true - Motorola manufactured the chips for a while, don't know about the whole time, but IBM was definitely there doing the Atari Jaguar.

  • No, no, this isn't a complaint about the moderation above or anything (Hey, I thought it was funny, too). It just struck me that this is about the 10th message in this list or so that was funny enough to be moderated "funny", several of which have hit the 4-5 point level. Just caught me off guard, that's all - at first glance this seemed like a relatively "dry" topic...

    I wonder if someone could get a paper puplished in some sociology journal somewhere on the correlation of topics and the types of comments (as measured by moderation rates) they tend to attract...

    Hey, Jon Katz, want to do a scientific paper? :-)

  • :P

    True, though marketing people will usually garner the big first sales. They will get the "look at this! it's cool" sales. R&D takes and says "you've used the cool stuff... now wouldn't it be even cooler if it could do *this* *this* and *this*?" so they get the secondary sales.

    IBM is definately a company that can wait, be the 2nd or 3rd or 4th one into the game and *then* whoops some ass with products other companies have marketted to hell so the public knows about them, and knows what would make them better (IBM's r&d)
  • by Pope ( 17780 ) on Wednesday June 13, 2001 @12:35PM (#154715)
    Why does this take three years?

    Allow me to read from the History of Video Games poster on my wall (USA release dates):
    Atari: 2600 = 1977; 5200 = 1982; 7800 = 1987.
    Nintendo: NES = 1985; SNES = 1991; N64 = 1996.
    Sony: Playstation = 1995; PS2 = 2000.

    Do you notice anything?
    I'd be *greatly* surprised if a console didn't take 3-5 years between generations. The legions of Sony fans who cried "it (the PS2) better be backwards-compatible!" just left me shaking my head in disbelief.
  • See their upcoming protein folder [], which should be able to do petaflops. Finally, chip manufacturers are starting to put some memory on the die (in the case of Blue Gene, quite a bit).

    Will game box manufacturers do the same? And how would it affect price if each chip had direct access to, say, 4-8MB on teh die? My guess is that price is what's holding it's cheaper and easier to produce memory that's 4 inches wide than 4mm wide.

  • You've got to consider though, that all of that money is going into a single project. The chip is only going to be used in the PS3...not sold to hundreds of PC manufacturers.
  • by Rhinobird ( 151521 ) on Wednesday June 13, 2001 @12:40PM (#154718) Homepage

    from the article:
    The three companies aim to design a "super-computer on a chip" with a wide variety of consumer applications, they said in a joint statement.

    "The result will be consumer devices that are more powerful than IBM's Deep Blue super-computer, operate at low power and access the broadband internet at ultra-high speeds," the statement added.

    Must resist urge...must...resist...must...res...


    please shoot me. for the love of god, shoot me...

  • IIRC the idea of the PS2 was enough to make people hold off on buying the Saturn and wait for the PS2 to come out, so they wouldnt have to buy another system shortly after they got the newest top-of-the-line machine.

    Maybe the idea of the PS3 will be enough to discourage other console makers from creating a new next-generation console. A company scrambling to put together a rival for PS2 will be way behind the game when it comes to making a competitor for PS3.

    I agree with you though, that most people probably wont wait 3 years for a new machine.. that's a long time.
  • Potential? Oh god yes. Start thinking Diamond Age. No, seriously. The BBC article only hits on a small proportion of the real story here. At the bottom of the VoodooExtreme page, some kind soul gave a link to this EETimes story:

    In short: by 2004/2005, IBM, Sony, and Toshiba will have developed the Cell processor, fabbed at .10 micron and running at 1 teraflops. It's fully scalable and will form the backbone for a true broadband internet.

    Neal Stephenson's science fiction does seem to be self-fullfilling, doesn't it?

    Ken Kutaragi speaks of this massively parallel beast in biological terms, alludes to peer to peer-ness on several levels, and says the OS that will run on the processors will be Linux-like.

    Now if that's not a made-for-Slashdot story, I don't know what is.

    The Playstation 3 is never directly mentioned, but given the very similar BBC story, we can assume that the PS3 will be one of the Cell processor's main uses.

  • IBM
    "The result will be consumer devices that are more powerful than IBM's Deep Blue super-computer, operate at low power and access the broadband internet at ultra-high speeds ,"

    Intel® NetBurst(TM) micro-architecture features
    400 MHz system bus
    Hyper-pipelined technology
    Rapid execution engine
    Execution trace cache
    Advanced transfer cache
    Advanced dynamic execution
    Enhanced floating point/multimedia
    Streaming SIMD extensions 2

    sounds like IBM and Intel share the same marketing/engineering deparments already

    Sean Chatman® with e-(ultra * advanced) pipeline double decker bus with streaming real-time kung fu grip execution algorithms!!!!!!!

  • For some reason VoodooExtreme posts a link to a BBC article dated March 12th, this is not news!

  • > I'd be willing to say this person has not a clue.

    You can be willing. You can say it. You can even believe it. But you'd have a hard time being more wrong.

    I've worked on projects with 100x the complexity of a game-box that came in under 3 years. And I have worked on bus design, bus-bridge design, processor design, processor validation, and the CAD SW to do all of it from colored blocks to partitioning to fault propagation to AMP test engines (ever run your combos and randoms off a beowulf cluster? I may have prior art in my name for that).

    That's where I get my perspective. I have a hard time finding computational devices I don't know all about. But until I bought my PS2, I hadn't owned a dedicated TV-game box. Ever. Not even an Atari. And I still don't, because apparently this is a game box/CD player/DVD player/network game console/web brower/interactive-tv terminal/surveillance device/base(are belong to us, all your, one each). I.e., it looks like a PC with some re-/de-featuring to make it feasible.

    I like the other answers people gave. Sony needs to milk the game developers. Platform stability, even if it's artifical, is the best way to do that. By constantly saying "3 years", Sony gives the game guys a hard point against which to schedule TTM and breakeven.

    Some people talked about the learning curve, but that's a simple problem to solve. Just document the thing better. Sony doesn't have to because it knows it has 3 years for people to read the release notes. I bet, though, there's some third pary out there who has written the PS2 equivalent of the Lions book and is making the SW world's life a lot easier.

  • This is off-topic, but you brought it up, so I'll respond. Excuse me, but I don't think a very high-quality and feature-packed iMac for $899 is a lot. And I don't think an incredible G4 starting at $1699 is so expensive. G4s are in short-supply, yes, but also keep in mind, Apple doesn't always necessarily have to pay more. Their contract (the AIM Alliance) guarantees a certain yield and a certain number of chips -- so if anything, Motorola is under pressure to perform, not saying, "Haha, pay more Apple!"
  • Actually... I've heard that matsushita supplies Nintendo with their mini-DVD drives "for free", while matsushita can use the GameCube design "for free" in their DVD-players. Apperently matsushita also owns some patent on the mini-dvd discs, so both nintendo and matsushita makes money on each sold game copy.
  • Probably a lot cheaper now, then.
  • The Nintendo64 used a R3400 and the PSX used a R3300 (I think, unsure about that one) so you might consider this synergy to have already taken place :-)

    Personally I think it's a good thing for consoles to use the same CPU architecture because it makes the compiler vendors' lives easier, which means their tools are better, which means game programmers' lives are easier, which hopefully means better games.
  • by Psmylie ( 169236 ) on Wednesday June 13, 2001 @01:44PM (#154728) Homepage
    Too bad it's not Intel. I was actually looking forward to a combination of the two commercials' sounds at the end... Imagine if you will:
    "Bling bling bling BLING - Play-sta-tion"
    It's got a certain rhythm
  • by Anonymous Coward
    After 14 years [] IBM has made the successor to the PS\2.
    I wonder if it'll have excessive features like more than 640K RAM...


  • The result will be consumer devices that are more powerful than IBM's Deep Blue super-computer, operate at low power and access the broadband internet at ultra-high speeds," the statement added.

    Does that mean I can invite Vladimir Kramnik (the current reigning chess champion) over to my house and have him lose against my Playstation?


  • Is Irn-Bru anything like the Cuban Iron Beer?
  • ...Platform stability, even if it's artificial...
    Listen closely; I work at a game company. Platform stability is *not* artificial.

    Firstly, how often has your PS2 crashed? Never? Exactly. If you have all the experience you claim, then you understand how long proper testing takes; you know the characteristic curve of bugfixes and how long you have to push it to promise an embedded device won't crash. This takes substatially more time than patching together a PC from spare parts at home.

    Secondly. Consistency. If you spend 18 months developing a game, and then it doesn't run on PS2(rev. 2002) or PS2(rev. 2000) because of subtle differences between versions, then you, your customer, and your publishers are going to be, umm, unpleasant to the manufacturer. Am I joking? Full of shit? Well, let's look to Microsoft for an example of the best consistency that multiple versions can offer: Win95, Win95A, Win95(OSR2), Win98, Win98SE, Win2000, Win2000Pro, WinME, WinCE, etc. A PC developer has to test on each and every one of them. You can understand how that might tax tester resources, why is probably why most PC games release several multi-megabyte patches to bring the game into a usable state.
    Incidentally, the list of Windows revisions I listed all happened in the time period between PSX and PS2. Also incidentally, any PSX game will play on any PSX from any year. Get the difference?

    Thirdly, development target consistency. Go buy the original playstation and a swath of games from 1995 to 2001. You'll notice a substantial improvement in quality and complexity. Console programmers get very close to the machine, and over time they learn to write with the machine's idioms, rather than generic approaches from, eg, school, Lions, Stevens, etc. This effect is not from lack of documentation -- even the manufacturer's games follow this rule. Nor is it from programmers taking "3 years ... to read the release notes.", and if I may stray into incivility, your are an ass for suggesting that it is so.

    Good day.

  • ... that the PC was by design (even if by accident of design) more amenable to hacking. I accept the criticism of my original comment.
  • Actually, from what I understand, IBM was going to fab Transmeta's chips. Transmeta ended up backing out of the deal, reputedly because IBM's services got to be too pricey.

    IBM does great work in the semiconductor business, it just takes a lot more than what Transmeta was willing to invest.


    [root@kgutwin /dos]# file msdos.sys

  • The N64 was easily 3 times as powerful as the playstation 1, it never cost 3 times as much. N64 was much more powerful, it did very well, but not as good as the playstation because of cartridges and the fact that it took an SGI to do development.
  • Ironic that the fastest PPC is from MOTOROLA not IBM (733MHz 7450 vs 700MHz 750CXe). By the way, that didn't split the group apart. It was IBM's refusal to adopt a SIMD.
  • Actually, I think it *is* the hardware that makes the system, and I think history bears it out. Let's see:

    Atari invents the home Pong console, practically the first of its kind.

    Other companies build various other boxes, some of which can play more than one game. The original Pong systems were outclassed, but then--

    The Atari 2600 comes on the scene and beats out the Odyssey, the Vectrex, and other consoles, mostly because of better hardware design. The cartridge format allowed for more and better games, and the hardware made better games possible. Therefore the software developers came, because the platform was powerful and relatively easy to write for.

    Nintendo comes along, and no one in this country had ever heard of them. Yet their console had such technical superiority and performance that the 2600 was doomed, and game makers sold their souls into Nintendo slavery just to get the right to code for it.

    Other systems, such as Sega's SMS, had good hardware, but Nintendo had already captured the mindset of the market and most of the software makers. There does come a point at which it's too late, despite technical merits, but it's technical merits which turn the tide.

    That's why, fast forward to the days of the PS versus N64, the PS won. The Nintendo had a better processor, but overall the PS was more advanced--it had CDs instead of cartridges. Game developers loved the CDs and hated the cartridges. To program a game into the small space of a N64 cartridge took more effort than if you have a full 650MB at your fingertips. Also, you don't need to license a proprietary cartridge format. So ultimately it was the superiority of the hardware which won over the game developers.

    I think this is going to be more and more the case, since there have been no major advances in gameplay for years. What is going to get more important is photorealism, and the platform which can offer the best realism and still have ease of programming on its side will win the developers come the next generation of consoles.

    We see the same thing in the PC gaming community, with video cards. 3DFX got game developers to code to their proprietary library because the performance was so much better and the effects the game devs could create were so much more intense. That abrubtly fell off when nVidia started making cards that were as good that people started buying, so the developers dropped the 3DFX-only route in favor of DirectX and OpenGL, which can be used with any card. But there was a period when 3DFX was so superior that gamers were only buying Voodoo cards, and so it made sense to code to the 3DFX cards only.

  • Uh...this was in Wired a couple issues ago. The PS3 in its current state is 16 PS2 boards on top of a high bandwidth crossbar. They want to keep the number of individual CPUs but don't want to have 16 PCBs in a single box. IBM as far as I know is the only company right now that has the ability to get four or more processor cores on a single die. From the stuff I've read I think the PS3 will really be a techno wonder, much moreso than even the PS2. The little game consoles we've been buying for the past couple years have been rendering engines primarily and thinking machines secondly. When you've got 16 processors on a high bandwidth crossbar you've got huge potential for not only media but also for logic and physics. Rendering millions of textured polys is a nice but its even nicer when those millions of polys make up a realistic looking cloth swaying, feathers of a bird buffeting in the wind, or gravel realistically being kicked up behind a car.
  • and while they talked about the ps2 just after the ps1 came out (1995 ps1, not that new thing), they mentioned a road map for the playstation series which consisted of Playstation 2 for 2000 and Playstation 3 for 2005. They also mentioned at the time the projected processing power they wanted to have in each of those. While they didn't know what ps2 would be, they knew they wanted roughly a 100 fold increase in power for the ps3. Looks like it's coming true. This is hardly new news. It's just finally coming to light and entering preproduction.
  • A couple years ago, so IBM could be making PS2 (I have seen people render Playstation 2 as PS/2) CPUs.

    The PS2 needs an adapter to use PS/2 mice and keyboards with it.
  • Densi wrote:
    >Blair wrote:
    >>Platform stability, even if it's artificial... >Listen closely; I work at a game company.

    O yay.

    (Sort of. I keep sending my resi to LucasArts and they keep saying "please do not include pictures of the Light Saber you constructed in your application.")

    >Platform stability is *not* artificial.

    By "platform stability" I didn't mean reliability. I meant compatibility. Minimal drift. To support the consistency you spoke of...twice. So that companies with shoestrings as skinny as WhizGameSchtremeCo's can get off the ground and pump a few more royal pct's into SNE's P&L.

    But, as long as you brought it up, yes, my PS2 has "crashed". Changing disks doesn't reset it. I have to hit the reset button after I insert a disk. You will now likely claim that's not a bug, it's a feature. Sophistry's a bitch.

    And yes, I do have experience with DO-178B Level-A certification testing of Avionics software. There is no more thorough test requirement. As a good buddy of mine puts it, "there isn't enough money printed to certify a TCP/IP stack for flight" (note: this is for cockpit/control systems; the rules are different for cabin/passenger-use systems).

    I attribute the fact that Sony was able to make the thing reliable at all with any sort of feasible business plan to the inference that the system is much simpler than some people are making it out to be. This can also be accomplished by building the complex system out of less-complex, well-validated subsystems. I note that my PS2 shipped with two quenched USB ports and a fat bay for an "expansion module". I.e., a year or so of buyer-beta testing will validate the main unit before it is integrated with these complicating components.

    >This effect is not from lack of documentation -- even the manufacturer's games follow this rule.

    Documentation is communication from one person to many. (Yeah, yeah, sometimes several people have to add pages to the doc; but still, one teacher per meme, many learners per meme.) Even the game developers within Sony have to read about what the hardware and system software team created. You do know they're not the same people, right? Thorough, accurate, readable documentation costs a lot of money. If the system is documented properly and the game teams are experienced professionals from theme to design to release, you would not see asymptotic improvements. You would see the best features of the system used well in the first games and any games thereafter. Cf. the difference between a detailed, precise, accurate map and "it's over there a few hundred miles".

    > and if I may stray into incivility, your are an ass for suggesting that it is so

    An mine ass were less civil and justified a like retort.

  • Nintendo64 used a custom designed R4300i, not a R3400.

Last yeer I kudn't spel Engineer. Now I are won.