PowerPC. A coup that became a fiasco. Will it degenerate into farce as the future of Apple looks bleaker than ever? Timothy Prickett analyzes the current players and their strategies. From Multimedia Futures, a sister publication.

Like everybody else that staked money on the PowerPC, Martin Burns, manager for Motorola’s European operations, is concerned about the desktop market in general and Apple in particular.

I think PowerPC is pretty well positioned for network computers, he says, and any desktop that doesn’t come from Intel. As for Apple, he says that Motorola will do all it can to help Apple sell its own and clone MacOS, and he is adamant that Rhapsody, Apple’s forthcoming operating system based on the NeXtStep object-oriented environment, will run on PowerPC chips. Apple will never move to Intel. PowerPC still has the best price/performance and multimedia performance in the market. If anything, Rhapsody will couple Apple more closely to PowerPC,he said. However, not everyone shares his point of view. NeXt runs on Intel chips, not PowerPC, and Apple is not committed, long-term, to the PowerPC, declares Stephen Smith, an analyst and managing director at Paine Webber. IBM promised Apple that it could deliver twice the performance of Intel chips and it dropped the ball. PowerPC has been a technical failure as well as a marketing failure. Even so, why would Apple risk a chip change at this stage and put Rhapsody on Intel chips? According to Smith, the move would be less disruptive than many might assume. 70% of the current installed Macintosh base is still using the Motorola 68K chip, he says. And, of the remaining 30%, the majority of those customers are still running 68K programs emulated on the PowerPC platform. Frankly, PowerPC is a niche player, says Christopher Goodhue, research director of end user computing at Gartner Group. Goodhue expects IBM to sell a fair number of its forthcoming G3 PowerPC Unix workstations. He says that IBM will have to push its new Windows NT Pentium Pro workstations a lot harder, adding that The volumes are going to be from users who want to spend less money than they would on Unix workstation – or users who want to run NT applications. Goodhue believes that IBM can make some headway, but he is less optimistic about Apple’s prospects with the PowerPC machines. The Mac market share of desktops is shrinking, although the Mac clone market is growing. Apple’s whole strategy lacks coherence and credibility. Copland didn’t play out, and we have to wait two more years for Rhapsody. Even where Apple has the advantage – a better fit and finish – Wintel is good enough for 95% of the market. In 1997, being different is not being better. Lamar Potts, vice president of operating system and technology licensing at Apple Computer, believes that the situation is not as dire as the industry watchers are suggesting. He argues there is no confusion about Apple’s support for the PowerPC.

The PowerPC is our platform of choice… and we plan to use it up to 2000, with more general plans beyond that. Rhapsody, he says, will run on the PowerPC, which does not preclude Apple from examining other processors. As to the immediate issue of selling as many Macs as possible – Potts adds: Apple has to turn the Macintosh around. For one thing, it is hard for Apple to invest in moving Mac technology forward if we don’t get the units up. In the wake of Heidi Roizen’s departure – she is the the Silicon Valley hotshot hired only last year to cultivate independent software developers for the Mac – Apple is particularly sensitive about its software problems. Software developers write programs based on seats and driving up that number will make independent software vendors more enthusiastic about MacOS, says Potts. He points out that Apple is making moves to address the needs of personal computer buyers, who are looking at features and price just as much as they are software. According to Apple and Motorola Computer Group – it sells embedded systems, clone Macs and motherboards to Apple clone suppliers – key to the Mac revival, will be the Common Hardware Reference Platform (CHRP). Aka, the PowerPC Platform.

CHRP to the rescue

CHRP is a set of hardware and software standards that enables Mac clones to use common personal computer components and be compatible with real Macs. The original intent was to support AIX, as well as MacOS and Windows NT on the CHRP, but NT on the PowerPC is finished and AIX is deemed inappropriate for most customers. The CHRP is expected to be finalised in May. Both Apple and Motorola think this will be instrumental in helping establish a viable and large Macintosh clone market. This will be the first time people can really make clone Macs, says Potts. Apple has only a dozen MacOS licensees. Intel has somewhere between 8,000 and 12,000 partners pushing its products. If we can get a fraction of these vendors to go with MacOS and CHRP, we will get the Mac base to grow. Apple says that it predicts the number of MacOS licensees will be in the hundreds by the end of this year, and as a bare minimum, it expects the clone market to triple from 1996’s figures of 200,000 unit shipments. Those numbers sound low, even to the Motorola Computer Group, which sold 44,000 of its StarMax Macintosh clones in the three weeks following announcement last November. It is still selling high-end Mac clones at that clip. The Group not only sells Macintoshes, but like IBM, it is permitted to sub-license the Mac operating system out to other parties. According to Dennis Saloky, marketing director at Motorola Computer Group, the company has four MacOS sub-licensees today and expects to double that number this year, and again the year after. He says that the Group wants to build a $1 billion systems business by the year 2000, and that CHRP, which will cut at least 10% out of the cost of building a Mac clone, will go a long way toward bringing Macintoshes into parity with Windows.

Full Circle

Not everyone is convinced that the CHRP will save MacOS and PowerPC on the desktop. The future of PowerPC on the desktop is tied to Apple, and that is not a pretty picture, says Nathan Brookwood, principal analyst for microprocessors at Dataquest. Apple has not convinced me that it will grow the Mac market, which we see flat this year and shrinking next year. As far as I can tell, Apple is not growing the base, it’s just shifting where Mac buyers get their computers. Everybody’s eating at the same table. Brookwood is convinced that the Mac base will continue to dwindle, but he holds out one hope that for Apple that may give it enough time and money to get its Rhapsody plans in order. Some people believe that a Mac upgrade cycle could help prop up sales for Apple in the near-term. But even with that, Brookwood says he is still pessimistic about the PowerPC’s prospects in bringing in new customers or attracting software developers. The Mac clone business is different from the Intel clone business, he says. The Wintel market is driven by Microsoft and Intel, and no one is worried about their long-term prospects. If an Acer or a Gateway 2000 evaporates, it has very little effect on the personal computer market as a whole. But if Apple goes down, the whole Mac clone business goes down with it. The difference is extremely important. Dataquest is equally unimpressed with the record of IBM’s PowerPC efforts. It doesn’t have a good high-end PowerPC workstation, and at the low-end, Sun is running roughshod over IBM, says Brookwood. IBM’s PowerPC server business is a good one, but Unix is a stagnant market. Most of the growth in the midrange will be driven by Windows NT, which PowerPC no longer supports. Many argue that Unix variants such as AIX are better operating systems than NT, but Microsoft has the resources to change that, and it will. None of this means that either IBM or Motorola will abandon PowerPC. I don’t expect PowerPC to go away any time soon, says Linley Gwennap, editor of The Microprocessor Report. IBM and Motorola will continue to invest heavily in PowerPC chips, and for the most part they seem to be keeping pace with Intel in performance. IBM is very committed to PowerPC for both the RS/6000 and the AS/400. It will continue to invest regardless of Apple. So where does all that leave PowerPC? In just about the same place that the Motorola 68K was during the years before the PowerPC was announced. Before 1991, the 68K chip was used as the main brain in Apple computers, and it is likely that 68K circuits were the basis of the CMOS processors in IBM’s AS/400 and 9370 mainframe systems. Motorola still has a very strong embedded processor market and is looking to adopt faster technology. The real difference this time around is that Apple is much weaker, and that can be attributed to the unremitting efforts of Intel and Microsoft. If IBM and Motorola had pipped Intel to the post, launching speedy chips, the odds are that Apple was still likely to stumble with Copland and fail to foster a lively clone market. All the excitement over the PowerPC effort and the fight for desktop control has amounted to nothing much. The Intel-Microsoft tag team has taken out the IBM-Apple-Motorola triumvirate without any real blood being spilt.*