The latest round of layoffs and restructuring at Apple may finally force the company to focus on the core Mac hardware and operating system products that are its key strengths, resulting in a strong Mac platform that Apple can ride to a return of its glory days. Certainly, the Mac licensees are betting that this will happen, thanks to the confluence of two events:
- The Common Hardware Reference Platform (CHRP) OS is due shortly, and the CHRP hardware designs are almost complete- Apple's refocus and downsizing mean it is more open to sharing the burden of success with othersThe three big Mac makers after Apple - Power Computing, Motorola Computer Group and Umax Computer - are gearing up their production facilities and sales and support staff in anticipation of significantly increased Mac sales. Power Computing and Motorola have recently doubled their production-line capacity, and Umax is poised to ramp up its capacity as well. In 18 months, the three companies could together deliver as many Macs as Apple does today - four million a year, versus their current estimated 400,000. With that kind of production, the pressure is off Apple to cover and succeed at every aspect of the Mac market - it can even retrench for a couple of years if necessary and know that the Mac platform won't retrench along with it.
The CHRP Foundation
The planned June release - after a year's delay caused by the Copland OS fiasco - of Mac OS 7.6 for the Common Hardware Reference Platform, followed by a Mac OS 8 release that includes CHRP compatibility a couple of months later, will give both Apple and the licensees the hardware-software base to truly move the Mac platform forward.
For example, CHRP systems will have system buses at 66MHz (and likely to be faster), compared to the 40MHz to 50MHz speed in most of today's Macs. That will boost the performance of today's PowerPC CPUs considerably, since the higher the ratio between the CPU and the system bus, the less efficiently the CPU can operate and the slower the system as a whole operates. For example, a 200MHz PowerPC 604e running on a 50MHz (a ratio of 4:1) bus is about 15 per cent slower than the same 200MHz 604e running on a 66MHz bus (a ratio of 3:1). On PCs, ratios of 3:1 are common, while on the Mac, ratios are usually 4:1, 4.5:1, or 5:1. When you realise that a Power Mac today is faster than a same-megahertz Pentium PC and then add in the boost of CHRP's faster bus, you'll see that the Mac can truly start to pull out way ahead of PCs in performance. And with the forthcoming 533MHz X704 Exponential CPU and 300MHz and faster PowerPC 604e and G3 CPUs from IBM and Motorola, that faster CHRP bus will provide the base on which to go even further.
Keeping costs down
CHRP also lets all Mac makers, including Apple, use less-expensive, more common parts, which should reduce Mac costs even further than they have dropped recently, and minimise the chances of product shortages that have plagued Apple in the past.
The first version of the CHRP OS is not an earth-shattering release - it adds support for PC-style serial ports, but is otherwise the same as today's Mac OS 7.6. The real value is that it divorces in almost every respect the Mac OS from the Mac hardware. In the past, Apple's OS has been tied directly into each model of Mac systems, requiring lots of extra OS code in ROM and making it difficult for both Apple and third-party developers to avoid bugs. The hardware-specific OS of the past also made it hard for Apple to use industry-standard parts, and it made it impossible for Mac OS licensees to create Macs from scratch - instead, they had to rely on Apple to provide the basic technology and the product leadership. No more: now each company can add value as they see fit, at the pace they see fit.
Apple is still central to the CHRP OS, so the Mac licensees are still somewhat constrained. For example, because the CHRP OS doesn't support the PC's ISA slot, Mac makers can't use it to provide a cheap, simple way for users to add internal modems to their Macs. Apple's Comm slot, for which Global Village makes internal modems, isn't part of the CHRP platform, and since the Comm slot is used on just a few Apple and clone models, no other company has developed modems for it.
Similarly, the CHRP OS does not support parallel ports, which would let Mac owners choose from a wealth of PC peripherals including the new Universal Serial Bus, which will likely become the standard input-device port for next-generation PCs; PC Card slots, for use in notebooks; power management, also critical for notebook designs; and the Firewire high-speed SCSI replacement interface, which is starting to find its way into digital cameras and other multimedia devices that would be natural for Macs to support. Apple needs to move on adding such support to the CHRP OS, even if that means doing so before Apple's own Mac designs support such slots and ports.
IBM Microelectronics (which will not make systems itself but has a CHRP design it is promoting to others), Motorola, Umax and Power Computing already have CHRP hardware designs ready to go or close to ready, awaiting only the CHRP OS from Apple. And big companies like Taiwan's Tatung are ready to make CHRP systems in volume. With players like these, the Macintosh can really grow.
The need to partner
Apple has yet to detail how its layoffs will affect its Mac OS and Mac hardware teams, but indications are that these aspects of Apple's business will retain the highest level of resources possible. That's essential to the Mac market rebounding in a big way. Already, it's apparent that the clone makers have helped reverse the Mac's declining market share in the overall computer market; even as Apple has had trouble delivering PowerBooks and non-Performa Macs, the clone makers have filled the demand, letting the overall Mac market actually grow for a change. Also essential is that Apple stop trying to go it alone. It needs to partner actively with its licensees, letting them contribute some of the missing components of the CHRP OS (at least two have offered to do so already).
Apple started this partnering when it formed the AIM alliance with IBM and Motorola to create the CHRP specification. However, the AIM alliance has been slow to act, taking several years to make CHRP a reality when it could have dedicated real priority to the effort and had it done much sooner. That slowness still exists: the CHRP specification effort for Universal Serial Bus, Firewire, and mobile computing are months out, if not a year. And that's just the specification, not the actual product design work.
Similarly, Apple has had an awkward relationship with the Mac licensees. After Apple decided to open up the Mac platform to authorised clone makers two years ago, it got cold feet and tried to pull back. Fortunately, the public outcry, led by Macworld, helped prevent that pullback. Apple then tried to prevent IBM from making Mac systems on non-Apple designs, using startup software from the Swiss company Quix.
More recently, Mac OS licensees have complained about foot-dragging on Apple qualification of their systems; about discussions at Apple that could lead to higher licensing fees for the fast, high-margin systems that Apple apparently didn't expect the clone makers to compete against it with; about slow availability of system software updates; and about uncertainties whether their Mac OS licences include the Mac OS 8 version due in July (they do, it appears) and the Rhapsody OS based on Next's OpenStep (Apple still won't give a clear indication of how Rhapsody fits into the licensing scheme).
The advent of strong Mac makers using their own designs is scary to some people at Apple. Scary because it raises the spectre of possible incompatibilities as clone makers differentiate their products. Scary because it raises the possibility that Apple could be shunted to the side by aggressive companies that outsell it at a time when it needs every dollar it can make.
In recent weeks, Apple executives have reiterated their support for Mac OS licensing. Those statements have taken the edge off some of the fears that the licensees have had about Apple's commitment to an open Mac market.
I hope that Apple executives see that an open Mac market is critical to the Mac's long-term success, and ultimately to Apple's as well. Apple has learned the hard way that it cannot do everything - the latest reorganisation and layoffs are a striking admission of that fact.
Apple needs to now face its fears and wholeheartedly support an open Mac platform. Work with the clone makers to ensure that the Mac platform as a whole succeeds in a way that supports Apple's ability to devote the right resources to the Mac OS - that's what Apple needs to do. A proactive, aggressive approach should give Apple the success it seeks with less of the burden to shoulder.