« The Sameness Engine, or, Write What You Know | Main | Let Us Not Forget the Original Allure »

Serial Missteps on the Parallel Road

Knowledge, once gained, is impossible to ignore, and so it’s difficult to remember what things were really like in the past. The attitudes and assumptions we held before time torpedoes them tend to disappear, and everyone acts as if they’d known the truth of the matter all along. But rewind to the year 2005 and recall that gamers, technology journalists, soccer moms and developers alike imagined the PlayStation 3 as a mysterious and formidable box with potentially earth-shaking powers. The Xbox 360 was a fine machine, the thinking went, but the PlayStation 3, when it arrived, would blow everything else out of the water.

I too became excited for the PlayStation 3 when the first rumblings and rumors about the successor to PlayStation 2, which had dominated the industry during its lifetime, came my way. The core of the project started with the idea, simple and true enough, that going parallel was the next big thing in computing architecture. Sony, Toshiba and IBM had been collaborating on the development of something called the Cell microprocessor architecture for years, and it was enshrouded by an unusually thick aura of novelty and exoticness. The net effect was that the PlayStation 3 became associated with– even synonymous with, in some of the more hyperbolic press– the entire future of technology.

So when the massive, ugly metal boxes with the hex readout and toggle switch on the front that comprised the first PlayStation 3 development kits began arriving at the studio where I worked, there was a charged anticipation in the air. Here we were, about to take a peek at the real capabilities of this much talked-about machine. As we stood around watching, a programmer quickly got an included demonstration program up and running. It displayed what was essentially a handful of flat-shaded cubes. We paused, a little confused– that’s it? The programmer quickly found the line of code responsible for the number of cubes, bumped it up to a thousand, and ran it again to see what would happen. The system halted in its tracks.

Over the next year, the team charged with PlayStation 3 development struggled valiantly against daunting headwinds. It’s commonly alleged that the system’s unusual architecture makes it more difficult to work with than most other consoles, but even if we had employed all of the best programmers in the world, we couldn’t have solved the other huge problem: the fact that the PlayStation 3 just wasn’t ready. Most of the way through development it wasn’t a system at all, but a processor on a circuit board that you could make do things. Until the very last second, nobody knew how its online components would work or what the platform requirements might be– Sony simply hadn’t written them yet.

The fundamental nonexistence of the PlayStation 3 as a real platform and game development system didn’t stop the powerful marketing organization behind the PlayStation brand from going into overdrive. If you flew to Los Angeles to attend 2005’s Electronic Entertainment Expo, there was a sticker on the television of your hotel room: “Welcome Chang3,” it read, as did numerous signs and billboards around the convention center. Expectations ran impossibly high, but that year, the press conference didn’t just meet them: it exceeded them beyond imagination. Trailers for upcoming games showcased improbable details like mud splattered on a windshield being smeared by a running wiper, or a close-up of steam rising from the sweat-sheened skin of a martial artist. Talk was breathlessly reported about what games might be like at hundreds of frames per second or running on two 1080p monitors simultaneously.

Sony’s booth on the show floor had a theatre, and attendees lined up to be admitted inside, to be subjected to a brief movie about “connected entertainment” with some vague, hand-wavy video sequences showing things like car shopping in an augmented reality where real-time computer graphics could be superimposed over and integrated into a live video stream. What that had to do with anything was anybody’s guess, but it hardly mattered. As people exited the theatre they filtered past a mockup of the PlayStation 3 enclosure, in silver, and a strange controller that looked like a boomerang. Everyone held up their cel phones to take a picture. “Come on, people, it’s just the box!” someone shouted.

One day, deep in crunch, I finally realized that PlayStation 3 games would be about as good as Xbox 360 games, in the grand scheme of things– that there were more similarities than differences in the two consoles’ relative power for typical video game software. If I was more technical, this might have registered sooner, but as it was, it took my actually working with the thing to understand it was as much a product of the current generation as anything else– not a piece of impossibly advanced technology that had traveled backwards in time to get to us. But most of the people I spoke to (who, to be fair, weren’t trying to ship a PlayStation 3 launch title) were taken aback by this notion. Such was the hold that the PlayStation 3 had taken of their imaginations that even claiming it would only be slightly better than the Xbox 360 was a shock to their understanding of the world.

It’s remarkable how long it takes for such perceptions, once fixed, to fade in peoples’ minds, however. The first cracks in the façade were exposed to the public at the following E3, in 2006, with the infamous press conference that saw the system’s specifications downgraded, its price revealed to be astronomical, and a flop of a game demonstration that had people snickering at the phrase “giant enemy crab” for months to come. But even after all that, and again after the system actually came out and you could inspect its games for yourself, people were still claiming that the watershed moment was yet nigh– just as soon as those lazy developers got around to “figuring out” how to unlock that inscrutable Cell’s latent superpowers.

* * *

The lessons from Sony’s great misadventure, of course, are well-documented: that greater raw strength does not mean greater fitness for the task at hand, that creating a development environment is not the same thing as creating a piece of hardware, that marketing needs to act in harmony with the actual product and set expectations responsibly. It’s too bad that the cost of learning them was falling from being the unassailable leader to third place among three. And the further unfortunate thing is that the incredible, often heroic efforts of those at Sony and elsewhere to bootstrap a next-generation entertainment platform into existence from nothing under great duress deserve much praise. This stuff is very difficult to do at all, let alone get right, and the final version of the PlayStation 3 ultimately delivered to consumers is by all accounts a superbly engineered piece of consumer electronics. I mean, hey, at least the GPU doesn’t melt off after a little normal use.


Comments (9)

Stern:

I don't think anyone familiar with microprocessor technology was surprised about the reality of the PS3. The technology for the hyped product simply did not exist.

By the way, are the initial development systems (still) covered by NDAs? I find it interesting how systems develop from first specs through successive iterations of development kits to the final hardware. Eg. in some interview Kutaragi said they initially planned to have a second Cell instead of a conventional GPU, but went with nVidia when they realized it wouldn't work.

Low-level engineers will probably tell you the RSX feels somewhat “tacked on” in the overall design, and that is almost certainly due to its late addition. I’m no expert, but I imagine Sony’s hardware engineers were slow to recognize the shader-centric world that video game graphics were inexorably moving towards at the time. I shudder to think how much worse off the PS3 would have been with two Cells and no RSX.

Merus:

To bring in the last of the three not mentioned in this vignette - I suspect that Nintendo's complete inability to rally developers around their current generation console is entirely due to a failure to build a development environment. Generally, it looks like it's felt (a "I don't know what I'm talking about and yet I'm going to do it anyway" phrase if ever there was one) that Nintendo's 'out to get' developers on its platform, something that isn't nearly as prevalent on Microsoft's side despite Microsoft actually being out to get certain titles.

The difference is, of course, that Microsoft's chief advantage is that they've nailed building a development ecosystem. They are incredibly good at it. It's what cemented Windows as the de facto standard of PCs, it's what made DirectX the de facto standard for PC games, and it's likely what will win them market share in this generation of consoles. Because Microsoft's good at making developers feel welcome on their platform, even if Microsoft are planning to steal their market, developers are a lot more comfortable working on it. I don't think Nintendo does that, and so despite having a massive audience (while Sony do of course have a much larger reach with the PS2) there's a huge reluctance amongst developers and publishers to jump on board.

Some Guy:

I remember not caring about any of the next-gen systems in 2005.

And I remember the PS2 being a wild banshee to develop on in 2000, just as the PS3 is a wild banshee to develop on today.

The Dark Prince:

So many people cite this idea that the PS2 was so hard to develop for but got good games in the end as a reason for future amazement in the PS3. It's a lovely argument that is flawed since the PS3 is not the PS2, it's just as likely to just continue not having good games because nobody will develop for it.

And as for engineering, well, the xbox has had a lot more thought in planning in the earlier stages and citing a problem in first generation models as a sign of bad design is hardly fair.

What I would like to know is why the PS3 is only "a little better" than the Xbox at best if it was a whole year later to market. Surely that extra time bought them something more than just not having a GPU that melted off the board?

"Developers, developers, developers, developers!"

As hokey and seemingly insincere as that phrase seemed when we first heard it, it turns out that Microsoft really meant it. It's been my experience that games are made on the 360, then "ported" to the PS3 at the last minute by a small team. Either that or the people making content on a day to day basis are doing it on the 360 as a team of PS3 programmers assimilates all the content. The 360 is just easier to develop on because the tools are so much better (and actually exist!).

Also, as soon as I saw the boomerang controller I knew the PS3 wasn't the next coming of god.

Neil:

First off, good article I enjoyed it. But you did say that "marketing needs to act in harmony with the actual product and set expectations responsibly." I agree that Sony did not, but neither did Microsoft by shipping a console that was defective hardware and died on almost a third of its customers, to this day. I mean that seems a little more important. I think that if people actually stopped buying Xbox b/c of that and Sony was leading now you would have written a different article. But alas, that is not the case. I just want to be told the truth, or at least something remotely close to it, by either company. C ya.

James:

I think the problem with the 360 is that Microsoft decided to one-up Sony on price and succeeded to a degree; however, in the process they forgot the cardinal rule of storage: It's _far_ better to have it and not need it than need it and not have it, especially because it's price drops exponentially. The other rule they forgot is that latency is a HUGE deal. There are no other comparable trade offs you can make in a computer that cut your latency more than replacing an optical media drive with a hard/flash drive.

Post a comment

(Comments are screened manually and may take a few hours to show up on the site. Thanks for your patience.)

About

This page contains a single entry from the blog posted on December 13, 2008 2:54 PM.

The previous post in this blog was The Sameness Engine, or, Write What You Know.

The next post in this blog is Let Us Not Forget the Original Allure.

Many more can be found on the main index page or by looking through the archives.

Contents © 2006, 2007, 2008 and 2009 Magical Wasteland. All rights reserved. The opinions expressed on this site are personal, and are not those of any company or organization with which the author may have an affiliation.