End of the duopoly?

Microsoft and Intel's informal alliance, known as the WinTel duopoly, has directed the development of PC technology.

The strategy of the two players--Microsoft, which controls the user interface, and Intel, which controls the hardware specifications--has paid rich dividends for the two companies, especially during the last few years as PC growth has skyrocketed. Both companies now claim more than 80 percent market share in their leading products--Intel with computational microprocessors, and Microsoft with PC operating systems.

As we enter the 21st century, it is obvious that the two companies are looking to reduce their reliance on each other, and are looking at other segments of the technology sector in order to continue their growth-and for good reason. The core PC market growth rates have slowed, and penetration rates have increased globally as we creep toward the mark of 100 million PCs sold for 1999.

Both companies have identified enterprise computing as the next segment that promises the greatest dividends in the near-term.

Historically, this is the segment that the WinTel juggernaut has been unable to break through-the domain of Unix. This situation is changing, however, with both companies' recent product offerings-Microsoft's Windows NT, targeted at enterprise, and Intel's Xeon processor, targeted at workstation and server markets.

Yet as the high-margin, high-end of enterprise computing becomes more and more visible, we may see that the alliance is starting to show some cracks.

Neither Intel nor Microsoft currently have any 64-bit offerings, which are crucial for the fault-tolerant enterprise computing segment. However, it seems neither party is willing to wait patiently for the other's technology to get up to speed.

Compaq, having gained Digital's Alpha and its 64-bit operating systems, is aiming to control the enterprise computing platform. Compaq is also collaborating with Microsoft to develop 64-bit NT to run on Alpha, while continuing to develop its own Unix offerings. Meanwhile, Silicon Graphics has announced its intention to manufacture NT-based systems.

Intel is also mounting its own assault on the enterprise. The Pentium II Xeon is its entr?e into the workgroup/mid-range enterprise market, following with Merced, the IA-64 offering expected in 2000. Intel is assidously drumming up software support for Merced and IA-64. Hewlett-Packard is a development partner for Merced, and we expect to see HP-Unix running on Merced.

Sun has also announced plans to port its own version of Unix to Merced. Intel recently announced an alliance with IBM, Sequent and SCO, a leading Unix-on-Intel vendor, in order to develop Unix for high-volume markets that run on IA-32, IA-64 and PowerPC.

We believe Intel is trying to create an industry infrastructure with extensive software support from ISVs for a Unix on Merced and later versions of IA-64.

Another area of divergence has been the embedded segment. Microsoft introduced its CE operating system as a platform for emerging information appliances, handheld PCs, palmtops, set-top boxes, and other digital consumer devices. Microsoft has been ahead of Intel in this segment, yet outside of OEMs that are doing their best to exclude the Microsoft "Trojan horse."

We believe that we will see Microsoft emerge as an important player in the information appliance segment. Intel has been a latecomer to the party, leaving MIPS, ARM and SuperH among others to battle out processor design. With StrongARM, Intel is in a much better position to attack this market-but we will have to wait and see how Intel decides to pursue this opportunity.

All in all, with the loosening of the WinTel alliance, the technology market could present a much more varied landscape going forward.