Uplink 4

“The Playstation 3 Issue”

March 2006

Contents:

• Introduction: The Next-Gen is Now

• Enter the Playstation 3

• Cell and the Geopolitics of the Chip Biz

• SNAP, Crackle, Pop: Nokia Takes on Mobile Gaming

Introduction

The next-gen consoles of tomorrow – Microsoft’s Xbox360, Sony’s Playstation 3 and Nintendo’s Revolution (a.k.a. the 360, PS3 and NR, respectively) – are today. The 360 is already arriving in stores, while the PS3 and the NR are slated to show up later this year. While console transitions are a staple feature of the videogame culture, this particular transition is much more than just another hardware upgrade.

What is at stake is a momentous geopolitical shift. Console hardware is no longer a US or Japanese monopoly, while game software is no longer the sole province of Nintendo and a handful of US firms. The next-gen consoles were designed and manufactured by globe-spanning networks of engineers, researchers and technicians. Console software is being produced everywhere from South Korea to Finland. Meanwhile, world-class videogame artists flourish everywhere from Kamloops, Canada to Mumbai, India.

It’s widely known that between 2000 and 2005, world videogame sales equaled and eventually surpassed the world box office receipts of film. This is only half the story, however. The past five years also laid the groundwork for what might be termed next-generation media infrastructures – specifically, high-definition digital broadcasting, widescreen TV, high-speed Internet access, and handheld or mobile gaming. The next-gen consoles are the devices which are going to pull all those infrastructures together, in ways we we can only begin to imagine.

The three articles in this issue of Uplink examine the PS3’s unique role in the latest console cycle, the arrival of the Cell chip and the geopolitics of multicore chip technology, and the world of cellphone gaming.

Enter the Playstation 3

To paraphrase Andy Warhol, the contemporary mass media could be defined as the perennial pursuit of fifteen seconds of fame. One of the most intriguing paradoxes of the videogame culture is the fact that it runs on much slower clock. The life-span of the average videogame console is measured in four to six year periods, called “console cycles”, while creating a quality videogame can take up to three years. To understand why the PS3 is such a breakthrough, one has to understand the crucial role these cycles play within the videogame culture specifically, as well as their gravitational influence on the media culture generally.

Console cycles have existed in every mass media – just think of the transition from black and white TV sets to color TV, or from analog broadcasting to high-definition TV. But if you’ve ever wondered why manufacturers don’t constantly issue upgraded videogame consoles, the answer is simple: it can’t be done. First of all, it takes two to three years for game programmers to optimize the software for any given hardware platform and unlock its true power. Second, by producing the same hardware over several years, console manufacturers can dramatically reduce their costs. The result is that consumers can spend $300 on a console for the sort of world-class gaming experience which would normally require a $3,000 investment in a personal computer.

Successful consoles are an artful balance of economics of scale with technologies of scope. The previous console cycle, which ran from 1999 to 2005, is a case in point. The three main consoles on the market were Sony’s Playstation 2, Microsoft’s Xbox and Nintendo’s Gamecube. After a battle royale worthy of Kinji Fukasaku, the PS2 garnered 60% of world market share. (The Xbox ended with 30% and the Gamecube with 10%.) Although the PS2 was technically outclassed by its competitors, what tipped the scales was Sony’s marketing and media savvy. The PS2 could play DVDs, while the Gamecube could not. Sony also invested heavily to create a deep third-party developer network, something Microsoft did only belatedly.

Today, each company has returned to the next-gen fray. Each is banking on a core strength – Sony on its legendary design and media skills, Microsoft on its fading (though still vastly profitable) PC monopoly and huge cash reserves, and Nintendo on its superb handhelds and unparalleled software library. Inexplicably, while Sony and Nintendo have both learned key lessons from the previous console cycle, Microsoft seems bound and determined to repeat past mistakes.

Just to forestall any misunderstandings, this is not a critique of the 360, which is a superb platform in its own right. This is a critique of Microsoft’s lack of a videogame strategy, which is a local example of a much more general problem – the inability of Wall Street-style capitalism to compete head-on with East Asia’s keiretsu capitalism or Eurocapital.

The single most important lesson of the 1999-2005 cycle is that videogames are evolving beyond a single platform. The first sign of this development occurred in Japan, Sony’s home terrain, where total videogame sales declined by about 40% from their peak in 2001 to mid-2005.¹ This was not due to any putative collapse in the videogame market, because Japan still has the highest rate of per capita spending on videogames in the world (about twice as high as the US). What happened was a fundamental restructuring of the game market. Console prices fell dramatically, while the sales of ubiquitous mobile devices and software took off like a rocket.

What happened to the console industry echoes what happened to the arcade videogames of the early 1980s, and the PC game industry of the 1990s, when less powerful but more numerous and affordable platforms overwhelmed the market share of their larger, more expensive competitors. Today, the center of gravity of the videogame culture is shifting to mobile media devices, ranging from the Nintendo DS to cellphones. In the not so distant future, every computer sold in the world is going to be accompanied by at least ten consoles, and well over a hundred mobile devices.

This is good news for Nintendo, which has always been strong in the mobile sector and in game design, but a headache for console-dependent Microsoft, and a potential nightmare for front-runner Sony. Sony’s first response was to launch its own portable mobile device, the PSP. But at some point between 2002 and 2004, the far-sighted Ken Kutaragi and the other executives in Sony’s game division realized this wasn’t enough. Sony could not simply upgrade the PS2’s graphical capacities and hope the mobile juggernaut would pass them by.

Instead, the Playstation team did something surprising and even nervy. They redesigned the PS3 to accelerate that juggernaut. The result is not just another multimedia machine, it is a key infrastructure of the emergent information commons. Five key design features of the PS3 make this possible:

     1. Support for the Blu-Ray disc format, as well as all existing CD and DVD formats. This means users will have a vast library of recorded and digital media at their fingertips, while opening the door to affordable downloadable broadcasting, movies, games and music on demand.

     2. Support for two generations of high-definition digital TV: the current standard of 720p resolution, as well as the more advanced 1080p resolution.² (The higher resolution will extend the PS3’s shelf-life for an additional one or two years, because TVs capable of running 1080p won’t be mass-produced until 2008 and 2009).

     3. Support for universal broadband and wireless access to the Web. Sony has garnered a wealth of online experience from its Everquest and Star War franchises, and is reportedly launching an online service for PS3 users, similar to Microsoft’s excellent XboxLive.

     4. Support for Sony’s superbly-designed PSP handheld, which already functions as a media player as well as a gaming device.

     5. Full backwards-compatibility with the vast library of existing PS1 and PS2 games. This is important considering the enormous size of the installed base of PS2s and PSPs, and the implied support for the PS2 platform.

From the standpoint of the Wall Street rentiers, none of these moves make any sense. Blu-Ray technology is new and expensive, screens capable of displaying 1080p resolution are not even in production yet, the spread of broadband service has lagged in the US, handhelds are a fiercely contested and competitive market, and backwards compatibility cuts into the surplus-profits of planned obsolescence.

From the standpoint of where the videogame industry is headed in the next five years, though, Sony’s strategy makes perfect sense. The PS3 is meant to create a commons where all the major game and media platforms can converge. Rather than charging consumers a hefty markup for admission to that commons, Sony is essentially subsidizing its start-up costs. Sony is not doing this out of the goodness of its corporate heart, but out of the pragmatic understanding that the next-gen console cycle is going to be driven by the co-existence and intermingling of multiple gaming platforms.

Sony deftly timed the entrance of the PS3 to maximize the returns from their still-thriving PS2 business, using the window between the announcement of the PS3 in spring of 2005 and its arrival in stores sometime in 2006 to release its portable platform, the PSP. This kept the PS2’s enormous fanbase happy, while enlarging the potential audience for PS2 software. It also gave Blu-Ray and high-definition TV an extra year to become mainstream standards.

The onus is now on Nintendo and Microsoft to answer Sony’s challenge. Right now, Nintendo appears to be in the best position to weather the Blu-Ray storm. Staying true to its quirky company history, Nintendo is carving its own unique path to the next-gen world, by focusing on a unique controller and a slimmed-down console which maximizes its core strengths in handheld design, game-play and software.

The outcome is less clear for Microsoft. The 360 does not have Blu-Ray capability, nor does it have 1080p high-definition support. While Microsoft’s XboxLive service is excellent, it has no equivalent of media devices comparable to the Sony PSP or Nintendo DS. Microsoft also fudged on the issue of backwards compatibility, by making this dependent on a software downloading system. Worst of all, Microsoft was in such a hurry to launch the 360 that the first machines had serious heating problems. Hardware review site ArsTechnica has the best overview of the issue:

Overheating is going to be a big problem with these things, and you have to set up your system and home theater with heat dispersion in mind. You’ve got five or six feet of cable between the brick and the console. Use that to ensure the power brick is a good ways away from the system, make sure air can move freely, and that the vents aren’t blocked in the slightest. While this sort of thing is second nature to PC gamers, console people may not realize just how much of an issue heat is with the 360 and kill their system before they even know what hit them.

Of course, without a huge warning on the box or the instructions this is going to be a problem. Microsoft isn’t being very upfront about how to properly set up the system, and they should have taken steps from the design phase on to make sure the system has no problem pumping out all the heat the innards generate. The 360 runs hotter than any other console I’ve ever seen, hotter than my computer, and a touch hotter than the space heater I'm using to keep my feet warm right now. Why Microsoft isn’t being more aggressive with educating consumers of this problem is beyond me. Telling people to make sure to keep the unit cool is a lot less of a headache than forum post after forum post complaining about glitched up systems.³

This is a temporary problem, of course, because the die size of the chips inside the console will rapidly shrink, reducing its power consumption. Still, it’s hard not to conclude that Microsoft was more interested in scoring rhetorical points over Sony than in delivering a consumer-ready product. The decision to launch the 360 before the silicon was ready bespeaks a shocking degree of cultural provincialism. These are power and heat levels which would be tolerated only by profligate US consumers, not by energy-conscious East Asian or European audiences. (Predictably, the launch of the 360 failed abysmally in Japan, though European consumers have been a bit more forgiving.)

This is unfortunate, because the 360 is an excellent platform, quite capable of holding its own against the PS3. The biggest hurdle is Microsoft’s inconsistent third-party developer strategy. The classic Wintel strategy of churning out patch-as-you-go software and using Windows as the all-purpose club to demolish the competition will not work in the videogame market.

Gaming firms don’t need operating systems, they need well-designed, reliable and affordable software tools to access the full power of specialized hardware, as well as distribution channels which don’t penalize third-party developers. Fans need customized (and customizable) game-worlds, as well as high levels of service and support. Videogame consoles are not automatic monopolies; they are complex ecologies of public, private and community services.

The dilemma for Microsoft is that the core competencies for managing such ecologies cannot be purchased like off-the-rack clothing. They have to be built from scratch, in the context of a long-term strategy which respects videogames as an independent media, while acknowledging constant (and sometimes punishingly critical) feedback from fans and consumers. While Microsoft has begun to shift its videogame business in this direction, most notably with standard-setting initiatives such as DirectX and a more open attitude towards third-party development generally, there’s still a long way to go.

Ultimately, both Sony and Nintendo will find ways to flourish in the next-gen environment. What is not clear is if Microsoft has the will to reinvent itself to succeed in the game market. This is too bad, because while it would be nice if someone gave the PS3 a run for its money, this isn’t likely to happen in the current console cycle. We may have to wait for Samsung or a plucky Korean startup to enter the console market in 2010 for someone to truly test Sony’s mettle.

– DRR

Endnotes

1. These figures are from a report on the Japanese videogame industry, available on JETRO’s statistical site (Web: http://www.jetro.go.jp/en/market/trend/industrial/pdf/jem0508-2e.pdf). Note that the figures are a bit distorted, due to the mild deflation which gripped the Japanese economy during 1999-2004, which means that nominal price declines translate as slight real increases. Thanks to Greg Costikyan for pointing this link out.

2. These figures refer to the number of pixels or dots on a screen. The 720p standard refers to a screen containing 1280 horizontal by 720 vertical pixels (roughly 900,000 pixels), while the 1080p resolution refers to a screen with 1980 horizontal by 1020 vertical pixels (slightly more than 2 million pixels).

3. Arstechnica Reviews. Accessed February 1, 2006. Web: http://arstechnica.com/reviews/hardware/xbox360.ars/5

Cell and the Geopolitics of the Chip Biz

It’s been clear for some time that the center of gravity of the videogame industry is shifting towards portable platforms. Handhelds and cellphones now make up over a third of the videogame market, mirroring the shift in the computer industry away from desktops and towards laptops. One of the interesting and unforeseen consequences of this shift is that the next-gen consoles are taking on the role formerly played by personal computers – that is to say, they are becoming a locus of software and design innovations, roughly comparable to the role luxury cars have in pioneering sophisticated devices (fuel injection, anti-lock brakes, etc.) which eventually become standard issue items in mass-produced autos.

Perhaps the single most striking expression of this dialectic of niche-market innovation and mass-market proliferation is the Cell chip which powers the PS3. The Cell is an awesome beast, created by the collective engineering might of Toshiba, IBM and Sony. It is not so much a single chip as a family of chip designs, designed to accommodate the sophisticated media, broadband communication and real-time processing requirements of future media platforms. The main reason for the Cell’s creation was the urgent necessity to bypass Moore’s Law, a famous maxim of one of Intel’s founders, which states that the computing power of microprocessors will roughly double every year. (Historically, this has proved to be true).

Moore’s Law just wasn’t fast enough for Sony, for two reasons. First, software speeds always lag well behind advances in processor speed, due to sluggish peripherals, operating systems and memory subsystems. Second, real-time media processing requires gargantuan amounts of computing power. The personal computer market worked around this problem by developing plug-in graphics cards, which crunched the media numbers while the main processor handled everything else. Unfortunately, the graphics processing units or GPUs which power those cards are complex and expensive.

How to reconcile processing speed with affordability? The console manufacturers discovered an elegant solution: piggyback on two computer revolutions, at once. The first revolution is the rapid advance in GPU technology, which has made enormous strides since the mid-1990s. The technical consensus is that GPU speeds have more or less quadrupled each year, about twice as fast as Moore’s Law would otherwise imply. The second revolution is the shift towards parallel processing software and multiprocessor technology.

In the past, each personal computer had a single microprocessor, connected to a range of external devices. In the early 1990s, researchers discovered that instead of constantly upgrading that microprocessor, an expensive undertaking, you could hook up a network of slower, cheaper microprocessors, and parcel up a given computational task into manageable bits. The trick is to build memory and data schemes capable of handling vast streams of data simultaneously.

This is precisely what Cell was invented to do. In its original design, the Cell had a central processor equipped with eight synergistic co-processors. The final design of PS3 reduced this number to seven, in order to increase chip yields and manage the complexity of the design. This does not mean, as one might assume, that the PS3 is only seven times as powerful as a personal computer with a roughly equivalent microprocessor speed. Each Cell co-processor, as well as the console’s built-in memory, has been fine-tuned for media processing and streaming.

The result is a machine so fast, it boggles the mind. It’s worth remembering that in terms of computational power, the very first Playstations were decidedly inferior to the average personal computers of 1995. The PS2, Xbox and Gamecube were considerably more powerful than the average computers of 1999-2000, although less powerful than the computers of 2003. But the PS3 is an order of magnitude beyond anything the personal computer market will deliver for the next five years. It is capable of delivering almost two teraflops of potential power – the equivalent of a small supercomputer.

It’s worth emphasizing that the 360 is an equally impressive advance in technology. Though slightly less powerful than the PS3, its PC-centric design does make it slightly easier to code for. The 360 bundles three roughly equivalent main co-processors into a single unit, fewer than the PS3’s seven co-processors. Sony decided that even though the PS3’s added complexity means short-term headaches for coders and extra start-up costs, the extra power will help designers over the long term. (Remember, Sony needed a machine which would last until the 1080p televisions start arriving in 2008 and 2009.)

What is perhaps most striking about the next-gen consoles is not so much their minor technical differences, but their geopolitical similarities. Their circuitry and design embodies some of the deepest tendencies of the multinational forces of production, while their software and media forms are the product of a multinational division of labor. Unlike the personal computer market of the 1990s, which was dominated to an extraordinary degree by Microsoft and Intel, no single company, nation-state or even group of companies has a monopoly on videogames. In the case of the Cell, Sony brought its media and design savvy, Toshiba brought its process engineering, and IBM brought its Powerchip expertise to the table. Samsung is producing the high-speed memory used in the PS3, Infineon is producing key elements of the 360, while a raft of smaller Taiwanese, Chinese and European firms are providing parts and components.

What is true for console hardware applies to the software as well. The sheer complexity and power of the next-gen consoles has driven development costs sky-high. This has forced game firms and studios to do three things. First, they are adopting cheaper, more flexible open source tools to speed the development cycle. Second, they are licensing middleware products, such as Havok’s rag doll engine or Ageia’s physics engine (readers with a fast connection can download demos at http://www.ageia.com/physxInAction/demos.html). Third, they are diversifying from consoles into handhelds, mobiles and cellgames. Rockstar released Grand Theft Auto: Liberty City for the Sony PSP, while Konami released a handheld version of Metal Gear Solid.

This internal or platform-based expansion of the videogame market has a qualitative dimension as well. This is the emergence of a gaming commons – public spaces modeled on the Web forums, mod sites, independent developer networks and fan communities spawned by the personal computer games of the 1990s. In a phrase, the loose-knit computer communities of the 1990s are solidifying into the console communities of the 2000s. This is something Sony seems to have realized very early on, no doubt due to the experience of running its massive multiplayer online franchises, and it will be interesting to see how Sony manages the roll-out of full-spectrum online services (gaming, messaging, and media streaming) for the PS3 later this year.

– DRR

SNAP, Crackle, Pop: Nokia Takes on Mobile Gaming

While the Sony PSP and Nintendo’s redesigned DS fly off the shelves, the cellphone game market is quietly shifting into high gear. Screen Digest estimates that the mobile game market surged from 850 million EUR in 2004, to 1.7 billion EUR in 2005. The industry is still heavily concentrated in the East Asian market, especially places like South Korea and Japan, which collectively make up about half of the market (while reliable data are scarce, the consensus estimate is that the Korean market alone was worth 300 million EUR last year). However, other regions are catching up fast. Demand for game-capable cellphones has exploded in the EU and US, while China’s nascent mobile game market earned somewhere between 100 and 149 million EUR in 2005.¹

Watching Nokia maneuvering deftly in the slipstream of the mobile game market has been almost as much fun as watching Sony’s console division play the press coverage of the PS3 like a harp. (Skeptics are advised to sift through the press coverage of the PS3 during January and February of 2006. Sony had no product, no interviews, and no announcements, but still managed to generate at least one splashy PS3-related story a day through its PR office, selected industry analysts and developer sites). While analysts have generally scoffed at Nokia’s gaming handset, the N-Gage, as a second-rate platform, the reality is that Nokia is making most of the right moves in the mobile field.

The analysts forget that Nokia never intended to compete head-on with the Sony PSP or Nintendo DS. Nor was the N-Gage ever meant to be a universal, one-size-fits-all game platform. As anyone who knows anything about cellphones will tell you, there are so many handsets, telecom providers, and telephone service plans on the market, that no single design could ever satisfy all customers.

Rather, the N-Gage was meant to be a mass market laboratory, where Nokia could test cutting-edge mobile games (e.g. Pocket Kingdom as well as classics such as Civilization), while driving forward some of the key manufacturing technologies necessary for mobile gaming. Nokia also made an explicit commitment to not become the all-devouring Microsoft of the mobile game world, by following an open standards model (e.g. licensing its Series 60 smartphone platform to other firms) and investing in the collective infrastructure of mobile gaming.²

One excellent example of this commitment is Nokia’s decision to work closely with the Korean game industry all the way back in 2002, via a cooperative venture with the Korea Game Development and Promotion Institute (KGDI). South Korea has long had one of the most advanced cellphone networks in the world, as well as a thriving infrastructure of suppliers, studios and content providers. Many Koreans are already watching TV on their cellphones for as little as $14 per month, and Korea’s efficient and well-run developmental state is pouring resources into the animation and media sector. Also, South Korea has specialized in CDMA cellphone technology created by Qualcomm, rather than the GSM technology favored by Nokia (GSM is currently the largest worldwide standard, with about 70% of the market, while CDMA makes up about 15%).

Instead of seeking short-term profits or trying to oversell the GSM platform, Nokia took the long-term view. It invested heavily in CDMA-compatible phones, funded other aspects of mobile software development (e.g. the Symbian operating system for cellphones), and provided tools and support to mobile developers.³ More recently, Nokia has created the SNAP program, a set of tools designed to turn Sun’s Java platform into a viable mobile gaming platform.

While the jury is still out as to whether Java can deliver on its promise of write-once, run-anywhere functionality, Nokia has taken care to apply the lessons it learned from the N-Gage to its bourgeoning smartphone portfolio. One of the prime examples is its media-rich 8800 platform, which includes customized sound-effects by composer Ryuichi Sakamoto. Launched in mid-2005, over 2 million of these high-end phones have sold worldwide. Although it will take time for smartphone technology to migrate to the mass consumer level, it’s clear that the technological base for mobile gaming already exists. What is still needed are better development channels for game companies and independent artists, that is to say, consumer-friendly and affordable service infrastructures similar to what DoCoMo’s i-mode provides in Japan. Nokia has done much to make this possible, but the mobile industry in general – and the telecom providers, in particular – will have to do more.

– DRR

End Notes

1. The higher estimate is from The Red Herring, the lower is from DigiTimes. Accessed March 5, 2006. The Red Herring. March 5, 2006: http://www.redherring.com/article.aspx?a=15413. DigiTimes. March 5, 2006: http://www.digitimes.com/systems/a20060111AA072.html

2. See, for example, the website of the Open Mobile Alliance. Accessed March 7, 2006. Web: http://www.openmobilealliance.org/about_OMA/index.html

3. Accessed March 7, 2006. Web: http://press.nokia.com/PR/200512/1026453_5.html