The PC is not dead

Several factors are driving industry growth. First, computer hardware is back in demand following the Y2K corporate spending slowdown, which lasted from the third quarter last year through this spring.

Second, corporations need more powerful PCs to run Windows 2000 (Win2K), which launched earlier this year. I expect Win2K will see increasing traction in the market.

Third, growth for x86-based servers should continue to accelerate because of the compelling price-performance this offers relative to Unix systems, for example, especially now that Win2K is starting to launch.

Fourth, the growing popularity of downloading music and video content from the Internet has many consumers spending more for more powerful PCs with more memory and faster processors that can support streaming media.

Although the adoption of Win2K has started off slowly, it should pick up momentum through the rest of the year, especially following the release of the service pack this summer. I believe most corporate IT departments are spending 6 to 9 months evaluating Win2K, prior to their launching the software throughout their companies late this year or early next year.

Even though the Win2K software may be launched later, we are seeing corporate buyers buying hardware today that is capable of running Win2K tomorrow. As a result, hardware configurations already are increasing dramatically in both processor speed and memory configurations.

Average memory per PC is increasing toward 128MB. Most system administrators recommend at least 128MB of DRAM to support Win2K, with closer to 256MB recommended for optimal price to performance. I believe the average memory per system was about 75MB at the end of 1999.

I also see an increasing mix of servers and portable computers driving higher average selling prices, higher margins and greater profitability for the computer hardware companies. Growth in servers, which accounted for 3 percent of 1999 PC units, and portables, which represented 17 percent of units, are exceeding the growth of desktop systems. This should lead to a richer product mix and higher margins and earnings for the computer hardware companies because servers and portables have higher price points and higher margins than desktops.

Server growth is ultimately being driven by the build out of the infrastructure that supports the Internet. The attractiveness of the price-performance offering of x86 systems explains why this segment has been growing faster than Unix servers. I believe that x86 growth should continue to accelerate as both Win2K and Linux gain traction.

Itanium, Intel's first 64-bit processor architecture, is due out in the fourth quarter and this should also help invigorate server sales. With Win2K and Itanium both available, I see this as the most compelling catalyst for computer system upgrades in the last several years.

I believe that 2000 will be the year of the x86 server, much like 1999 was the year of the high-end Unix server and 1998 was the year of the sub-$1000 PC.

We are also seeing an interesting dynamic in portables. As the performance of microprocessors for portables is closing the gap with desktop microprocessors, and battery life also is improving, portables are starting to become a no-compromise replacement for desktops in the workplace. As an increasing percentage of the workforce becomes mobile, portables are becoming the computing platform of choice. I believe that this trend will also benefit the computer hardware companies because portables have higher selling prices and higher margins than desktop PCs.

Consumer PCs also are being sold for higher profits this year than last. Over 40 percent of all retail PCs being sold today have built-in rewriteable CD drives to take advantage of downloadable music on the Internet. Many PCs are being bought with software for image manipulation to use with digital cameras. Given that the sub-$1,000 PC phenomena started in 1997 and the average PC hardware replacement cycle is three years, many of those buyers are returning to get a more full-featured PC today.

All of these factors are combining to make 2000 the best year of revenue and EPS growth for computer hardware companies since 1996, in my view. I believe this sets up a fundamentally different environment than 1999 and I expect to see computer hardware stocks continue to outpace the market.