This year's Macworld convention in San Francisco showcased Apple Computer's recent metamorphosis. The new Apple is slick, sleek, savvy, and sexy--Silicon Valley meets Hollywood. Apple has gone from the brand of counterculture technological revolutionaries to the brand of the Beautiful People. Steve Jobs has discovered that in an industry characterized by high-volume, low-margin cutthroat competition, one can make a comfortable living creating a marquee brand affordable to those with a healthy amount of disposable income. Never mind that Apple has reduced prices in recent years; when mom-and-pop working-class America compares an Apple with a Dell, the latter usually gets the sale. Other than Apple's core constituency of professional graphic designers, musicians, and die-hard Mac-heads, Apple is left catering mostly to the hip and the cool (and those who aspire to such). To cement this image, Apple partnered with BMW, integrating the iPod music player with BMW stereo systems.
Thus situated at the high end, Apple now finds itself with approximately 2.5 percent of the personal computer market, down from 10 percent a few years ago. Despite this low market share, the company is profitable and enjoys good margins, having reduced production costs and keeping a low head count of talented employees.
Apple's computers have a lot going for them. Apple's new Mac OS X operating system is a well-done Unix-based hybrid that is a joy to behold and use. More importantly, it's largely immune to the security problems that are almost daily news in the Windows world. The slew of applications for Mac OS X are likewise very polished and user-friendly. This has been Apple's hallmark almost from its inception; it started with the Lisa, found acceptance in the first Macintosh, and has continued to the latest incarnation. Apple's Mac OS X is undoubtedly the major differentiator from the competition.
Apple's computers are also reasonably fast: Apple has continually switched microprocessors over the years in its quest for performance, and the recent batch of G5-based Macs have compared favorably with the competition when it comes to raw computing power. In addition to those selling points, recent Macs have captured imaginations with their physical style--sleek, sexy, somewhat industrial, exuding an enigmatic quality. Apple has complemented this with advertising campaigns appealing to the high-energy, intelligent, style-conscious, and somewhat iconoclastic crowd that buys the machines.
The big news at this year's Macworld was Apple's switch to yet another microprocessor, this time Intel's new Core Duo, which is basically an extension of Intel's successful Pentium M found in many laptop computers. This was a good move for Apple, but one that belies problems and appears to have created some new difficulty. The impetus for this change was stagnation of the Power 5 microprocessors supplied by IBM. The Power architecture had reached its limit (just like Intel's Pentium 4). Power 5 processors also generate high levels of heat--so much so that they are unusable in Apple's notebook computers. This has been a serious problem for Apple. True to form, it did something about it.
Apple also introduced new software at Macworld, most notably a program called Front Row. Front Row is a control panel for digital media that is accessible from a handheld remote controller. Front Row is the linchpin of Apple's plan to use Macintosh as the centerpiece of personal digital entertainment. Front Row integrates with iTunes, Apple's popular online digital music store. Front Row also lets users purchase and view videos and preview movie trailers online.
Looking beyond the superficial, Apple's switch to Core Duo hints at the challenges a company in its position faces. Although Apple engineers have long been involved in microprocessor design with business partners Motorola and IBM, the company has never produced its own. This leaves it dependent on others for the heart of its systems, a situation familiar to many in the computer business. Apple unquestionably had to progress from the Power processor, which leads to the question, Why Intel?
Enter IBM. IBM has quietly been developing a revolutionary new microprocessor, the Cell, with partners Sony and Toshiba. The Cell appears to be everything IBM has promised, offering a leap in performance--video performance in particular. The Cell has demonstrated the ability to process and output video data in real time. Sony's new PlayStation 3 gaming computer, due for retail release sometime, uses the Cell for its central processing unit. Marketing talk says that PlayStation 3 will be 50 times faster than PlayStation 2. It is undoubtedly significantly faster (more likely by orders of magnitude) than Apple's new Core Duo systems.
Why should Apple worry about a gaming console? For starters, adding a 2.5-inch hard drive, keyboard, and mouse to the PlayStation 3 turns it into a full-fledged personal computer. The real eye-opener is Sony choosing Linux as the operating system for PlayStation 3, breaking away from Microsoft. The PlayStation 3 should initially retail for around $500, an amazingly low amount considering the computer's staggering performance. The PlayStation 3 also enjoys stylish good looks, for which Apple also is famous. Indeed, the two companies have been fierce rivals for the past several years.
Apple appears intent on dethroning Sony as king of consumer electronics. In Sony, IBM has an eager partner, hungry for success against a strong rival. As a company, Sony is a huge multinational with an impeccable pedigree, a track record of success, and vast experience in consumer electronics. This brings back the question, Why Intel for Apple? Why is Apple not using the most revolutionary new microprocessor to come along in many years, one that takes video performance to the next level? The likely answer is, IBM never offered it. Regardless, this will prove to be a big problem for Apple.
Video performance is the lifeblood of those who use computers to process photos and other graphics for a living. Programs such as Adobe Photoshop have long been the killer apps keeping Apple's core customers in the fold. These applications are available for the Windows platform, but most users prefer the friendlier user interfaces found in Apple operating systems. If these applications were ported to Linux, allowing them to run on the Cell processor, graphic professionals would question whether to handicap themselves for the sake of a friendlier user interface. Faster rendering equates to more work performed in the same amount of time, which in turn fosters greater productivity.
Faster performance is necessary for electronic music professionals who need digital signal processing muscle. Those constituencies have always been the first to buy Apple's fastest computers, and Apple depends on the revenue stream they generate. This illustrates the importance of both hardware and software to certain market segments; the importance of the two aspects changes as hardware becomes more powerful. Software (such as Mac OS X) has usually been the differentiator, because no company has been able to offer hardware that performs compellingly faster than the others. The Cell processor promises to change that. There is a very real possibility the digital media market will quickly migrate to Cell-based computers, if the applications are available.
Demand for the Cell apparently is strong, even before release. The processor has many applications. Military system companies are lining up for it, due to the powerful video processing, number-crunching, and digital signal processing capabilities. The Cell may bring new possibilities to cellular phones, a potentially large market for the chip. Toshiba will use the Cell to power large-screen, high-definition televisions; Toshiba also produces laptop computers. Automotive companies are looking at the Cell; robotics will benefit from it; medical instrumentation companies are embracing it. At least one company is currently building a supercomputer from Cell processors.
Gaming will represent perhaps the largest market, along with general-purpose computing. Gaming is a market Apple hopes to gain share in; at Macworld, Apple presenters continually spoke about the gaming capabilities of the new Macs. Cell-based systems will preempt Apple's growth in the gaming market.
Don't overlook IBM's other large partner, the Chinese company Lenovo, to which IBM sold its entire personal computer division last year. Cell processors may eventually appear in Lenovo computers. The Lenovo deal displayed IBM's business acumen: if you can't beat them, join them. Rather than expend effort trying to compete with Lenovo and others in a highly protected Chinese marketplace, IBM instead sold the business to Lenovo, keeping an 18 percent stake in the company. Now IBM has the blessing of the Chinese government, enjoys rock-bottom production costs, and still gets to offer PCs in Western markets through co-marketing. IBM decided that 18 percent of profits from a huge enterprise is better than struggling for 100 percent of profits from a much smaller one.
An overlooked factor that may come into play is Linux. IBM's server business benefited by embracing Linux several years ago. Strong indications suggest it is about to push Linux onto the desktop. The Open Source Development Labs (OSDL) in Beaverton, Oregon, was founded several years ago to promote open source software and standards. IBM is one of the primary sponsors of OSDL, and in 2003 it scored a major coup when it hired Linux creator Linus Torvalds to head all research and development.
Torvalds' primary focus these days has been the development of a strong software infrastructure for what is known as the desktop environment, a term used to describe the graphical user interface and tools needed for the average, nontechnical computer user. Torvalds is in charge of OSDL's Desktop Linux initiative, which is scheduled to release a desktop environment this summer.
Though a very capable environment, the Linux desktop to date lacks the polish of Apple's Mac OS X. Many of the applications available under KDE and GNOME (the two major Linux desktops) are in various stages of development and do not compare favorably to their commercial counterparts. Some progress has been made recently. Maya, a popular professional rendering and animation program, has been available on Linux for some time. GIMP is a free program almost identical to Photoshop; it lacks the latest Photoshop features but is a very capable program. Many scientific researchers have discovered that the Linux version of Matlab, a popular mathematical program, performs faster than the Windows version. Several years ago, Adobe experimented with porting to Linux. Sun Microsystems' Star Office suite is a Microsoft Office look-alike that has received good reviews. The entire suite of KDE applications continues to make slow but steady progress. KDE also recognizes Macintosh keyboards; it can behave much like a Macintosh, with the proper configuration; and it offers Mac OS X-style menus, translucency, and mouse-over effects.
If OSDL's Desktop Linux initiative can produce something usable by the general public, then expect IBM to work with Lenovo to deliver a Linux desktop to at least business users. If the Linux desktop ever gets good enough for mass consumer use, then companies such as Sony and Lenovo can offer it and lower their production costs: companies currently pay around $100 in licensing to Microsoft for each copy of Windows that ships with a computer.
One of the elements Mac OS X and Linux have in common, which differentiates them from Windows, is their relative immunity to security problems. This is a strong selling point for the public and should contribute to wider adoption of Mac OS X and Linux. One thing seems certain: Linux usage is about to explode, thanks to the Sony PlayStation 3--analysts expect Sony to sell 200 million of them, which means 200 million Linux computers in the hands of mostly young people (between 5 and 18 years old). This is phenomenal, and is perhaps the best possible way to push Linux into the mainstream.
One criticism of Linux for mass-market use is the issue of device drivers for peripheral components, such as printers, video cards, and scanners. This has been a problem for the Linux community, which largely comprises technical users and hobbyists. Popular opinion claims that device manufacturers will never work closely with the Linux community, and thus Linux will not be viable in the mass market.
Companies like Sony and Lenovo have an advantage in dealing with this issue--the same advantage Apple enjoys by controlling all aspects of Macintosh production. Sony and Lenovo have a large degree of control over their products. They deal in economies of scale and can choose which components are used in their products. If they want to offer Linux compatibility, all they have to do is build systems standardized on components compatible with Linux. IBM has already proven this approach, ensuring that IBM ThinkPads run Linux well. Linux is also compatible with a wide range of peripherals, from printers to scanners and digital cameras. Optimal video performance is still problematic for Linux, but Sony is working closely with nVidia on the PlayStation 3 video controller.
Apple faces other challenges. Macworld magazine's recent independent benchmark testing of the new Macs concluded that the new desktop Macs are not as fast as Apple claims. (Benchmarks for the MacBook Pro laptops should be available after their release.)
Apple also released a Power PC emulator, called Rosetta, that allows Mac OS X programs written for the Power architecture to run on the new Core Duo systems. Benchmarks show some Photoshop functions run about 50 percent slower under Rosetta than they do natively; this makes Rosetta a utility unsuitable for most professional-level Photoshop work. For other applications, Rosetta results are mixed.
Apple has also provided a certification program, called Universal Binary, that software manufacturers can use to offer versions of their products guaranteed to run natively on both the Power and Core Duo architectures; they basically will provide two copies of a program in each box. Apple hopes independent software vendors will adopt this approach, to avoid angering customers who are not ready to buy the new computers but who need to buy software now and don't want to pay again when they step up to the new hardware. Adobe Creative Suite Production Studio (a software bundle of Photoshop and several other Adobe titles) sells for $1,700, so paying twice for it is an expensive proposition. That was a problem for Apple when it transitioned from OS 9 to Mac OS X, and it hopes to avoid similar trouble this time around.
So far, software vendors such as Adobe have balked at producing Universal Binary-certified applications for their latest versions. Doing so will incur additional expense and lead to inventory complications in the distribution channel. The software vendors are more likely to offer upgrade versions, a more palatable option for them. Whether future versions will be Universal Binary compliant is the subject of much speculation.
Apple's relationship with software developers has been a bit tense lately. Apple has been developing new applications that it either bundles with Mac OS X or sells separately. A photo-processing program called Aperture received prominent display at Macworld. Aperture is a powerful photo-management program that lets users easily organize large amounts of photographs. It also performs many of the basic photo editing functions found in programs like Adobe Photoshop. Apple describes Aperture as a program "Designed for Professional Photographers."
Feature for feature, Aperture doesn't compete with Photoshop directly, but Adobe has to be concerned with the direction Apple is headed. At Macworld, Adobe released the beta version of a program called Lightroom, which is essentially identical to Apple's Aperture. Did Aperture force Adobe to release a product still in beta stage, in an effort to salvage its development? Some of the new Apple software bundled with the Mac provides features that users previously had to purchase elsewhere.
Apple's new software strategy may have the independent software developers thinking about new ways to hedge. Of course, the best hedge would be to start offering versions for a higher-performance competitor, such as the Cell. The Cell also gives upstarts an opportunity to enter the graphics software market, usurping established players such as Adobe. If the established software companies don't port their products to Linux, IBM has the option of creating an emulator, such as Apple's Rosetta, to allow Mac OS X and Windows software to run on Linux. Given the Cell's quickness, software operated in emulation mode may perform at acceptable levels.
There is already a Windows emulator for Linux, called Wine (an acronym for Windows Emulator or Wine Is Not an Emulator, depending on whom you believe). Wine is far enough along in development to be finished and pressed into service. Early indicators imply many software companies are working on Cell ports or at least investigating the platform. Sony reports distribution of 4,000 Cell software development kits.
Apple's recent investment in retail outlets is similar to the move toward application software self-sufficiency. Apple now has 130 retail stores, many in upscale locations such as Tokyo's Ginza district and San Francisco's Union Square. This approach, while giving Apple more control (and potentially more profit), also incurs large overhead. Market saturation is a risk, or demand for Apple products may hit a threshold, with Apple then committed to maintaining the retail infrastructure. Apple's retail strategy requires it to maintain a certain level of continuous sales. Traditionally, Apple's reseller network has weathered periods of slow Apple sales by offering a varied selection of third-party products. The resellers also have latitude to experiment with new third-party products. Apple's new outlets have affected the resellers. Some of Apple's old-line resellers have gone out of business. Several have filed a class-action suit against Apple.
In what promises to be a challenging year for Apple, an additional complication has occurred. Walt Disney Co. has agreed to purchase Pixar Animation studio, a move that makes Steve Jobs Disney's largest shareholder and gives him a seat on Disney's board of directors. Jobs has been able to manage both Apple and Pixar. The question now is whether his new Disney role will prove a distraction from his work at Apple. Jobs is very image conscious, and many consider him a business icon. Now that he finds himself on a much grander stage, he may well become preoccupied with it. This possibility could impact Apple's success. The Disney relationship does offer some positives: Apple can offer Disney content--and co-branding possibilities now exist.
Ironically, the Cell processor should enjoy huge success with animation studios, as the new workhorse for their render farms.
To summarize Apple's current position: The company appears to have reached a point where it cannot reduce its operational costs. It has taken on the continuing fixed expense of a large retail infrastructure. Its new hardware architecture is not as powerful as first perceived, and migration to the new platform is likely to create some tension with existing customers and software partners. To maintain leadership in the digital media market, it has had to reduce prices of its iPod media players and will most likely have to continue to do so. Last, Apple will face a strong challenge from Cell-based systems from Sony and others.
Some have already taken notice of Apple's prospects: in December, financial analyst Andrew Neff of Bear Stearns lowered his rating of Apple stock from Buy to Hold.
Apple has made some moves to improve its position, at least as far as its iPod media players are concerned. By the end of this year, about 50 percent of all automobiles sold in the United States should offer iPod integration with their stereo system. Apple is also opening very small versions of its stores in mini-malls and other places, in an effort to reach out to more customers. While Wall Street may have concerns about Apple's profit potential, Apple does have enough margin cushion to lower computer prices, which should attract buyers to the retail stores. The question is, how will Apple's new business model measure up? Is it sustainable?
Apple's recent branding and retail strategy appears to be an attempt to create a "Nike effect." For years, teenagers from impoverished areas have been one of the largest consumer groups of expensive Nike athletic footwear, spending up to $200 per pair. Apple has achieved this effect with the iPod, but can it do it with its computers? One interesting recent phenomenon is an apparent increase of Mac Mini sales due to user desire to integrate the iPod with a media server.
What can Apple do to counter the coming IBM/Sony/Toshiba/Lenovo challenge? For starters, it can hope the Cell processor turns out to be overrated. This is a real possibility. The computer industry is littered with the remains of technology that promised greatness yet never delivered on the hyperbole. Barring that, Apple should start looking at other microprocessor technology. One great advantage of Unix is the ease of porting to different hardware architectures. Apple may need to migrate Mac OS X yet again in response to the Cell's challenge.
Several companies are working on technology similar to that of the Cell. Sun Microsystems has been developing an 8-core microprocessor, code-named Niagara, that may compare well with the Cell. Some startups are also working on high-performance microprocessors. Apple may have to choose one of these as a partner. Apple could do this and still continue producing systems with Intel's Core Duo, creating a dual-tier product line to satisfy both casual and power users. Sony and Lenovo will most likely adopt such an approach, at least initially until they see how the market for Cell develops. Another possibility is IBM deciding to offer the Cell to Apple. Finally, recent industry speculation is that companies such as Intel and AMD will simply copy IBM's Cell architecture; the feasibility of such an approach may well be decided in a court of law.
There are some who think Apple should allow Mac OS X to run on other hardware. This would be an interesting strategy. The main advantage would be a much larger market. Some disadvantages are vastly lower margins, having to support a plethora of third party hardware, and having to compete with low-cost vendors offering the same operating system. Apple's only differentiator would be the physical design of its computers (which could possibly work, if the "snob effect" came into play). Apple, when pressed to port Mac OS back in the mid-'80s, refused to do so. Now Apple is again in that position. What is the best course? It makes sense to port Mac OS X only if Apple's current strategy doesn't work. Again, a great advantage of Mac OS X is its underlying Unix architecture, which enables easy porting to other platforms, making it easy to execute a third-party porting strategy.
IBM faces its own set of challenges and risks. Foremost, overheating may still be a problem: the Cell's primary core is a 4GHz Power 5. IBM claims to have solved the problem with advanced on-die thermal management and improved physical characteristics. Market timing and coordination between IBM and its partners will be important. IBM is still dealing with a legal challenge to Linux, and though it appears to have won that battle, an even fiercer fight involving software patents may follow (in which IBM will most likely prevail, should it come to pass). Geo-political forces could interfere with the Lenovo partnership. Despite these potential risks, IBM appears to be in an enviable position.
Sony also faces some challenges. Rumors claim the PlayStation 3 will cost between $800 and $900 to produce; Sony plans to sell it for around $500. This is a common market strategy in the game console business. Manufacturers sell the consoles at a loss, and then recover the losses and proceed to profitability from sales of games and related products. Sony recently announced a delay of PlayStation 3 until November, allegedly because of copyright problems with the BluRay DVD Sony intends to use. Despite this, there is a chance Cell processors will find their way into desktop PCs later this year. IBM's initial system push for the Cell appears to be blade servers.
2006 promises to be a very interesting year for the computer industry. Sony and Apple are on a collision course in the coming battle for digital entertainment and media systems. Apple has the early edge, while Sony appears ready to start executing long-term plans. Significant new hardware and open source software introductions may change the very nature of the industry. How Apple and other companies cope with the coming changes will make for interesting observation. Ultimately, the consumer should benefit.
Adrien Lamothe developed his first computer program in 1976 and never looked back.
Return to ONLamp.com.
Copyright © 2009 O'Reilly Media, Inc.