The Cell Chip - how will MS and Intel face the music?

Part II: Place your bets

Meet the utility company

The Cell proposition, we now see, has a couple of appeals to an audience which appreciates that computing technology hasn't got very much better in ten years, only slightly cheaper. The respective marketing departments of IBM, Sony and Toshiba have decided not make these too explicit at this early stage.

The first is the economic appeal of outsourcing computing cycles, if not entire IT departments. If you think the United States or the EU would never outsource the production of such a vital resource to such an unknown, in a potential trouble region, then just think oil.

The Grid argument proposes that it doesn't really matter where your data centers are, and the Cell at home proposes that you don't need a PC to make apply computer processing to digital media. The Cell promises to make computing as we know it - both commercially, and privately, invisible. But can anyone claim to deliver on such bold propositions?

Let's look at the business computing proposition first.

The idea that computing power is a utility isn't new. Back in 1999 many companies were investing heavily in the idea. HP, Intel, IBM and Sun all made the same bet, but only IBM and Sun seem to have stayed the course. Intel's global data center project was ignominiously written off a couple of years ago. Now Sun's vision, based on Scott McNealy's corny car analogy, is all that's left to compete with Cell, and Sun explained some of the numbers this week. (Read this for more details). Sun has spent a lot of money to make Solaris as flexible and responsive to this utility model as it can, and we can now see this was a wise bet. Sun too has complex, multithreaded machines in the pipelines, but only Sun, or maybe Google, has an OS that can bounce processes around the world.

So many factors must be overcome for such a model, either on IBM's terms or Sun's terms, to succeed, that only a fool would predict the future. We know that business managers demand ever increasing returns on investment and that IT investment can bring genuine short-term competitive advantages. The utility idea turns the conventional IT department inside out, and IT managers must become cunning procurers of cheap cycles or simply bespoke software houses. Tending to the machines no longer brings a competitive advantage, the logic goes. But if there isn't a level playing field - a fair market for cycles, and open and equitable access to useful APIs and other software resources - then the utility model collapses. Hardware itself isn't enough to make utility computing come true.

Remember that the politics of utility companies are keenly fought over. The desire for speculators to make an Enron out of a simple utility (to overlay the abstract concept of "market" where "market" doesn't really need to exist) has not gone away. Utility computing isn't magic, and doesn't make such problems go away. Experience has taught us that people really seem to prefer local utility companies, because they're cheaper and more accountable.

So for now, grid computing looks like a theory in search of some evidence. A bit like a Cell looking for some hardware resources. It's a great theory - but we're not convinced people want it.

But that's nothing compared to the problems Cell will have in the home, among people who want to share the bits of popular culture they value.

We can digitize it for you wholesale

The internet may one day be seen as a short-lived quirk, a weird Channel 79 on cable TV. "Occasionally useful" sums up how many people see it today, quite reasonably, or at least the 50 per cent who sign up. There are enough jokes, novelties, prospects of dates, or enough free music and porn to make this hopelessly unreliable and whacky channel just about worth it. That's all though.

However the internet has undeniably contributed to an explosion in the distribution of culture product, and that's largely because the devices allowed us to do it. People value this end of the technology proposition very much indeed. It's the one aspect that exceeds the hype. But it only took place because copyrights holders couldn't restrict the exchange of these cultural goods. Technology pundits tend to see the internet and argue that it's changing society. In fact, the reverse is true. Internet technologies are just the latest in a long line of tools that we've adopted because they allow us to store and transmit our culture.

Such tensions betweeen rights holders and the public are not unique - and there's nothing unique about "digital culture" either. Copyright holders really aren't interested in stopping stuff moving about, they're simply interested in being paid. We've been here many times before. Radio at first presented a much more dramatic threat to song sheet salesmen and honky tonk piano players than the computers and internet present today, but we all got over it.

Each time, a social settlement has been reached that allowed people to recompense the industries who invested in those storage and transmission technologies. The same deal takes place over and over again. We're in a funny time at the moment, because that settlement hasn't been reached yet. But it will be, and once it is, the Cell vendors had better start thinking of a more attractive sales pitch than "stops you doing stuff."

"Let's you do stuff" will be a start.

We can begin to see how important Cell is, even if we can see how its designers ambitions will never be fulfilled. Already, we suspect, Cell is a catalyst that is sweeping a lot of half-baked business plans into the bin where they really belong.

So to return where we started: for Intel and Microsoft does indeed does a deep philosophical challenge. The answers each company must produce to counter Cell aren't technical - faster bits, swishier graphics - but relate to how they can engineer social settlements that benefit both themselves and their customers, and traditionally technology companies haven't been very good at these. Intel, at least, with its new appetite for lobbying, indicates that the future isn't entirely a technical challenge. At this early stage, only Intel is facing the future with one eye open. ®

Related story

The Cell chip - what it is, and why you should care

Digital settlements

Digital music: flat fee futures
How the music biz can live forever, get even richer, and be loved
Why wireless will end piracy and doom DRM and TCPA Jim Griffin


Similar topics


Send us news

Other stories you might like