Mac Death Match, Round One: Chaffin vs. Enderle
MacNewsWorld presents a rumble in the Silicon Jungle -- May 7th through May 14th -- a six-round Mac Death Match in which Mac Observer editor-in-chief Bryan Chaffin and the always-controversial industry analyst Rob Enderle square off on today's key Mac issues.
In today's media-drenched culture, few things generate more buzz than celebrity defendents, reality-show contestants and -- Macs.
Why does the Macintosh whip up such passion from both its adherents and detractors? How has Apple Computer, a company that produces just a miniscule percentage of the world's PCs, become such a cultural force? Why does the iPod Mini elicit the sorts of oohs, aahs and omigods customarily reserved for Brad Pitt and Jennifer Lopez?
Those questions may not be answered for generations. Nonetheless, MacNewsWorld seeks to examine these Mac-centric phenomena through its new opinions column, "Mac Death Match." In the tradition of CNN's Crossfire and MTV's Celebrity Death Match, Mac Death Match pits those on ridiculously opposite sides of the Mac debate to battle each other in print.
Each combatant is asked to opine on one Mac-related issue. After receiving both polemics, MacNewsWorld collates and returns both responses to each. Then these bitter rivals rip apart each other's assertions -- to the death.
Rob Enderle, a TechNewsWorld columnist, is the Principal Analyst for the Enderle Group, a consultancy that focuses on personal technology products and trends.
Bryan Chaffin is editor-in-chief of The Mac Observer, a leading online source for Mac news and information.
MacNewsWorld: Okay. IBM PowerPC or Intel?
ROB ENDERLE ENTERS THE RING
Apple's last financial report indicates its PC revenue is growing at five percent in a market that is growing between 16 percent and 20 percent, depending on whose numbers you trust and the region analyzed. The company steadily has been dropping share -- granted more slowly recently -- for the last decade. Much of what makes up its computers comes from others. The PowerPC comes from IBM; the OS kernel comes from FreeBSD, and other vendors manufacture much of what it sells. Very little Apple intellectual property (IP), as a percentage, makes up the Mac today.
However, Apple's physical design is market leading; the user experience is second-to-none; and user loyalty is off the chart. In addition, publications like Consumer Reports rank the user service experience best in class, while ranking the value of the products themselves in the middle to bottom of class.
With the iPod, Apple proved that, if it built something that mapped to the Windows standard, it could outsell everyone else. In that market, which grew a whopping 87 percent, Apple grew an incredible 900 percent. The difference: Apple focused on common standards. iTunes runs on Windows, and the iPod is plug-compatible with Windows desktops and laptops.
The Power of Standards
This shows the power of standards, market standards, with the two most powerful standards in the segment being Windows and x86. When Steve Jobs made his own choice, at NeXT, he chose Intel, and clearly x86 (often AMD x86 interestingly enough) is where the FreeBSD kernel is most common. Jobs certainly could have used Motorola or IBM back at NeXT, but he apparently realized that the market simply would not accept a new product that different. NeXT was still too different and didn't get off the ground, even though a lot of us lusted after NeXT machines.
IBM, as opposed to Motorola, does bring a lot to the table but even IBM's own desktop division doesn't use the PowerPC because it simply is not accepted by business as a standard, even with Linux or FreeBSD running on it. It is accepted with AIX (Advanced Interactive eXecutive, IBM's proprietary UNIX OS) but only in very small quantities and only in another declining market, the RISC workstation market, which has been bleeding share (look at Sun) to Intel, Windows and Linux for some time.
Window Closing Rapidly
Looking ahead Apple has two major problems on the horizon. Microsoft's "Longhorn" release, like Windows 95 in its day, is slated to close the gap dramatically. Granted, Longhorn really is targeting Linux, but it will do a lot of damage to the Mac OS in passing. Meanwhile, Linux itself, where firms like Novell and HP are leading a massive effort to build a Mac-like OS, has similar potential. Apple may not be the target of either effort, but, in what is likely to be the competition of the decade, Apple simply may be collateral damage in a market choosing between two platforms, neither of which is the Mac OS.
Apple's window to raise itself above one of these two competing platforms is closing rapidly. It does not want to be number three in a market that wants only two. And -- let's be clear -- what both Linux and Windows have in common is x86 at the core. Economies of scale continue to make the x86 platforms the most cost-effective. And competition between Intel, AMD and VIA continues to push cost, power and now security to ever more aggressive levels.
IBM is trying to push back, but IBM's own internal divisions are tied tightly to Intel. IBM Microelectronics is actually building AMD and VIA parts, making it increasingly difficult for them to justify a non-x86 effort. The division used a unique x86-based processor in its Xbox victory. This blend of chip technologies IBM owns and has licensed will add another variant to the mix and make the non-x86 products even harder to justify because of the massive cost pressures on that Microsoft deal.
In other words, even IBM may have trouble sustaining PowerPC two to three years from now in its current form. Clearly Motorola is all but out of this market for similar reasons. With SPARC and Alpha in solid decline, virtually every x86 technology looks more attractive than PowerPC long term.
Finally, one of the biggest emerging problems is showcasing PowerPC's advantages. As Popular Mechanics recently demonstrated in its April 2004 issue, benchmarks accepted on the x86 platform don't run on the PowerPC platform without significant modifications which made it look like Apple was cheating. In fact, Popular Mechanics basically said Apple was cheating, indicating there may be no performance advantage with the PowerPC today.
If the PowerPC costs more due to a lack of economies of scale; if porting costs for the FreeBSD core, application porting costs, motherboard integration costs and drivers than Intel; if third parties like Maximum PC and Popular Mechanics prove it underperforms x86; and if the platform is due to be eclipsed by x86 in all related markets, then Apple has no choice but to make the move.
The longer the company waits, the more expensive it will become.
BRYAN CHAFFIN'S OPENING PUNCH
For better or worse, Apple is married to the PowerPC platform. There are various and sundry reasons for this, but the primary reasons are the issues of legacy and entrenched development, the need to control the whole widget, and the real issue of processing power.
In reverse order, let's look at processing power. While Motorola's abandonment of real advancement in G4 development was a tremendous problem for Apple and one of the primary reasons the company lost market share in the last few years, the G5 is hands down the fastest processor family in the desktop space. MHz for MHz, the G5 stomps on Intel's processors, although comparisons to AMD are far closer.
The Power of PowerPC
This is particularly the case in Apple's stronghold markets of digital content creation. Better yet, IBM seems committed to increasing the MHz speed of the G5, something long gone at Motorola. That means that Apple won't have to be explaining the MHz myth, a very good thing when it comes to customer perceptions.
Moreover, the PowerPC line, including the G5, uses less power and puts out less heat than Intel's processors, making them ideal for products like the Xserve and the PowerBook line.
For example, Intel has had to emasculate its mobile line of processors to reduce heat and power consumption. In contrast, Apple uses normal G4 processors in its portable line. Certainly Intel's new Centrino line is an improvement in that company's mobile offerings, but I'll take a real G4 over a Centrino any day. Once IBM works out its G5 production issues, Apple will be sitting pretty, even in the MHz arena on which the public has been trained to focus.
Working the 'Whole Widget'
Many of the people who want Apple to move to Intel want to be able to run Mac OS X on their non-Apple hardware. Were that to happen, Apple would no longer control the whole widget, and the Mac platform would no longer "just work."
In addition, Apple would be faced with having to support what amounts to a near-infinite number of hardware configurations, sending the company's support costs soaring by many orders of magnitude. Look how well Microsoft manages to do -- or not do -- even though it spends more on research and development every year than Apple grosses. Apple couldn't begin to compete against Microsoft in this market, especially without the revenues it generates from hardware sales.
Of course, Apple could move to Intel processors and keep the hardware locked (meaning that you wouldn't be able to run OS X on anything other than Apple-branded Intel boxes), but how would this help the company? The company would be trading power, heat output and power consumption for what?
Honoring the Legacy
The biggest issue, of course, is the issue of legacy development. Mac software has been developed for the PowerPC, which includes extensive development of AltiVec-enhanced code (AltiVec might best be described as MMX on speed and steroids combined). Photoshop, Final Cut Pro, Shake, InDesign, QuarkXpress and even Mac OS X itself -- all of these vital apps make use, sometimes extensive use, of AltiVec. Were Apple to move to Intel, developers would have to abandon this area of development.
In addition, Mac developers would be forced to retool their development yet again, causing many to balk. Developers had to retool for the transition from 68k to PowerPC and then again from Classic to Mac OS X. If they were asked to change yet another time, Apple would be faced with the challenge of trying to get enough developers from the Windows world to "switch."
This, however, is where the kicker comes into play: Apple would be doing so with a fraction of the user base that is has now.
Should Apple move to Intel, this would result in millions of Macs already in use becoming legacy systems with no future. Apple would have to start from scratch in trying to rebuild its user base, and who is going to develop for that?
Apple could do several things to mitigate these factors. It could provide emulation layers for PowerPC systems that allowed them to run Mac OS X for x86 code and then another emulation layer that allows x86-powered Macs to run PowerPC code. But unlike the Classic Environment that allowed for such a smooth transition to Mac OS X 3 years ago, such emulation would face an enormous processor tax. Users would have a poor experience, which would translate into even poorer sales than Apple has today, with the possible exception of niche markets, such as servers, where these legacy issues are less relevant.
Check back May 10th for Round Two.