How Sun can Magic away Apple's consumer success

Analysis While everyone was talking this week about FreeMac.com's upcoming million-Mac giveaway, an announcement that will have far more impact on Apple seems to have passed them by. Sun Microsystems said it is working on a super-chip for Internet access devices. Dubbed the Microprocessor Architecture for Java Computing (MAJC -- or 'Magic' -- for short), Sun's processor design promises a major leap forward in processing power at a price point way below today's desktop CPUs. Of course, with FreeMac.com's offer due to kick off in less than a month's time -- Magic won't even enter volume production until next year -- and tying in nicely with the pundits' 'flavour of the month' topic, the free PC, it's not hard to see why it grabbed all the headlines. IT industry watchers have been keenly tracking the emergence of the free PC model ever since FreePC.com (no relation, apparently) began offering gratis computers in exchange for a multi-year commitment to its ISP partner and an agreement to be bombarded with online advertising. FreeMac.com's offer isn't even the first time a free-computer company has chosen to hand out Macs: New York ISP OneStop Communications did that back in February. Still, the free PC concept continues to generate plenty of interest among consumers, and that in turn is forcing the major PC vendors, including Apple for all Steve Jobs' dismissive MacWorld Expo comments, to take notice. Some manufacturers, such as Compaq, have even been sufficiently concerned by the trend -- though it's debatable whether half a dozen small PC vendors really constitute a trend -- to experiment with it themselves. The point is, the need not be worried by it, and they certainly shouldn't give the business credibility by getting in on the act themselves. In essence, the point is that free PCs can only makes sense as a business model if they're very, very cheap. The cheaper they are, the lower the margins the vendor makes on each box, and so the more ISP subs they have to sell to make the deal work. And the fact is, because these machines are PCs, there's only so far that manufacturers can go to bring down the cost of each unit. And if the cost of Internet access also shrinks, that makes it even harder to fund the purchase of the PCs you, as an ISP, intend to give away. So the free PCs is essentially no more a threat to Apple than any other sub-$500 box, which, once you've factored in the cost of a monitor and other essential add-ons, isn't that much cheaper than an iMac anyway. But what if all those cut-price Internet access boxes -- of whatever form, and whether they're free or not -- aren't PCs? Suddenly, the game changes, and this is why the Sun Magic announcement is important not the current crop of cheap and/or free computers aren't. Almost all computing devices currently designed for Internet access are, like the iMac, general purpose personal computers dusted down and marketed at Net-hungry consumers. This shouldn't surprise anyone, since PCs have to date been the best way of getting on line -- the Internet is, after all, just another network to hook your client computer up to. However, that's changing. Firstly, the perception that the Internet is just another computing application is breaking down. Instead, the Net is being transformed into a medium for new forms of commerce and new forms of communication -- it's gone way beyond the super LAN that it was developed to be. So it's no wonder that so many consumers are buying the iMac not as a personal productivity tool -- ie. a computer -- but simply as an Internet access device. At the same time, new kinds of device are being given network connectivity in general and Internet compatibility in particular, from the next generation of digital mobile phones to handheld gadgets like the Palm. The upshot of all this is that what matters is getting on the Net, rather than how you get online. If a $200 box connected to your TV provides the same access to email, QuickTime TV and Web-based homeshopping as a $1000 PC, why bother with the latter? Of course, plenty of users do need a general purpose PC as well as an Internet terminal, so there will always be room for the iMac in the market. However, consumer sales will fall back to where they were before the advent of the Internet, and for an increasingly consumer-oriented company like Apple, that could spell real trouble. Sun's Magic enters the picture by providing a very powerful yet cheap (company insiders reckon the chip will ship for around $50) platform for this new generation of Internet access devices -- aka iMac killers. Magic was designed as an Internet-oriented multimedia processor, handling streamed video and audio, 3D graphics, and communications data efficiently and quickly. In that respect, it follows the trend set by Intel with its MMX multimedia extensions and more recently by Motorola's AltiVec technology, set to debut real soon now in the PowerPC G4. The difference is that Magic has been developed afresh with features like these well integrated into the core, whereas older chips, including the G4, have had them bolted onto previous architectures. Magic has also been designed for multi-processing and will support, like the next-but-one generation of the PowerPC, multiple CPUs cores on a single chip. Magic will ultimately compete with Internet appliance reference designs from a wide range of hardware companies, but it has the advantage that it's based on technology designed for the job rather than older PC-oriented processors, and it can leverage all that Java programming experience out there -- though it's important to remember that Magic isn't a Java-only gig; it will also run compiled C and C++ code. It has to be said that Sun has yet to release full details of the MAJC system and what it can, in reality, do -- that's going to happen next week at the HotChips semiconductor industry conference. But if Magic delivers only half of what Sun is promising, it will still form the basis for some very smart Internet access devices -- devices, more to the point, more than capable of matching the performance of top-end consumer computers. And that could well make it very difficult for Apple to sell iMacs. So forget all this guff about free PCs -- the real challenge to Apple's pricing, hardware and consumer strategies will come in 18 months' time when the first Magic-based machines appear. ®