Netbooks

This week’s issue of The Economist has an article, “Small but Disruptive”, on the growing popularity of small, relatively simple portable computers, commonly called netbooks.  These machines buck the general trend of personal computing since its start, of ever faster, more powerful machines.  The trend has been reinforced by the falling cost of computing power, in accordance with Moore’s Law, and has been accompanied by the use of more and more powerful, and resource-hungry software.  This has suited two of the key players in the industry pretty well: Intel has been able to deliver increasingly poweful chip models, which can run more complex and feature-laden software from Microsoft; and demand for the software helps boost the market for new chips.

The traditional idea of a PC is embodied in the name, “personal computer”.  The PC’s initial appeal was, in large part, that it gave an individual a computing resource that was under his or her direct control, in contrast to the centralized computing services provided by, for example, a corporate IT department.   (When the IBM PC was first introduced, in 1982, the Internet, in anything like its current form, did not exist.)    Although the advent of better networking and the Internet itself changed this perceotion to a degree in some environments, where Sun’s slogan “The Network is the Computer” had some resonance, mostly the PC has still been thought of in a context of one user = one machine.

This view was not unanimous, of course.  There was an ongoing discussion about the possibility of developing a computing “appliance”¸ which would perform simple functions (for example, Web browsing as originally envisaged, or E-mail) with the ease of use of a television.  The WebTV device, introduced in 1996, which provided Web browsing using a television set as a display, was one attempt to provide such a product.  And there were some more mainstream products that were focused on compact size and portability, not just raw performance.  (I had an IBM Thinkpad that I bought in 1995, at a premium price, because it was much smaller than most laptop computers.  I kept, and used, that machine for 10+ years, despite its technical obsolesence, because it was easy to carry around, and allowed me to do the things I really wanted to do with a portable computer: E-mail¸ writing, and basic Web access.)   But these products were generally a bit outside the mainstream:

Simple, low-cost laptops are not a new idea. There were several abortive attempts to make them in the 1990s, but big computer-makers soon dropped them for fear that they might undermine sales of traditional laptops.

Another strand of development was the One Laptop per Child project, started at MIT by Prof. Nicholas Negroponte; the product objective was often referred to as the “$100 laptop”.  Although the project to date has not managed to drive the cost down to the $100 figure, it sparked considerable interest in the core idea of a simple, low-cost machine that could be used for education.  (I worked some with OLPC, and there were some quite creative ideas in system design and user-interface design that came out of it, especially since the target ‘”customer” for the laptop was a not-necessarily-literate child.  Readers who are interested in the project can have a look at the project Wiki; it contains more detail, and links to downloadable versions of the OLPC software.)

Asus, a PC maker in Taiwan, came out with a commercial product, the Asus Eee PC, which it started selling in late 2007, for $250.  The machine used solid-state flash memory instead of a disk, a seven-inch diagonal screen, and the free, open-source Linux operating system.  Despite being outside the mainstream, the machine was a hit:

By the end of 2008, Asus had sold nearly 5m Eees and all its rivals had jumped on the bandwagon. In total, 21m netbooks will ship this year, almost twice the number in 2008 and more than 15% of the entire market for laptops, according to Gartner, a market-research firm.

Along the way, netbooks with more standard configurations appeared.  For example, machines arre now offered that use the Microsoft Windows XP operating system.  But the price competition in the netbook market forced Microsoft, in turn, to make significant price concessions, which in turn have impacted profitability:

To keep Linux off most netbooks, Microsoft has had to discount Windows XP from $40-45 per copy to $10-15.

The appeal of the netbooks is, of course, considerably enhanced by the continuing development of what has been called “cloud computing”: services provided by centralized servers, and accessed ovder the Internet, using a Web browser.  (Google Mail and Documents, or Facebook, are examples.)   Intel has so far gotten a fair slice of the netbook market, in part because it introduced a new line of Atom low-power processors.  But Intel is not entirely above the fray, either:

Still, just as with Linux, there is a credible alternative to Atom: processors based on designs by ARM, a British firm, which already power most of the world’s mobile phones. Several Taiwanese PC-makers are working on a netbook-like gadget that would be powered by an ARM chip and run Android, a variety of Linux developed by Google for use on smart-phones.

This threat, in turn, may be producing some cracks in the Microsoft-Intel axis:

Intel is also pushing another version of Linux, called Moblin. Its long partnership with Microsoft notwithstanding, Intel does not seem to care what operating system netbooks use, provided they contain the firm’s chips.

The founder of Microsoft, Bill Gates, famously dismissed the Internet as “a fad” in the early 1990s.  Mr. Gates is a very smart man, and has long since realized he was gravely mistaken.  I can’t help but wonder if his recent withdrawal from day-to-day operations at Microsoft (where he was a famously detail-oriented manager) is his way of trying to get out at the top of his game.