CNET también está disponible en español.

Ir a español

Don't show this again

Tech Industry

Commentary: Digesting the server glut

According to Meta Group, recent reports of slowing server sales are actually a trailing indicator for the technology market.

    Recent reports of slowing server sales are actually a trailing indicator for the technology market.

    The leading indicator was the reduction in IT budgets and general delay in the release of new projects that has characterized user planning and spending for the last two quarters. We expect the server market slump to continue through the end of 2001 as the market digests the infrastructure growth of 1999-2000 and projects continue to be delayed.

    However, we predict that a combination of factors, including the maturing of new server technologies such as Intel's Itanium, new capacity demand and an uptick in spending on new projects in early 2001 will lead to renewed growth in the server market by mid-2002.

    Several market forces have contributed to slowing server sales. First, many of the dot-coms that purchased technology at a frantic pace through the summer of 2000 and helped drive Sun Microsystems' sales in particular have faded from the picture. Many of these companies are no longer in business, and their infrastructure--sometimes still in the box--flooded the secondary market in early 2001. Surviving companies are being much more cautious about spending.

    The telco industry, another major source of networking and server purchases, has also been hit hard by the economic slowdown. Finally, Web hosting, ISP and application hosting companies have also slowed or stopped spending as demand for their services wanes and they go through their own shakeout.

    Now, instead of investing tens of millions of dollars in building new data centers, companies that need additional capacity (or simply have cash available) can buy significant infrastructure for pennies on the dollar, as witnessed by the recent Digital Island acquisition.

    Compounding the slowdown in new projects and the glut of product on the market is the fact that many organizations overbought in 2000 and now have 30 percent to 40 percent of their server capacity available for expansion. These companies rightly view new server purchases as discretionary enhancements that can be delayed. We expect these companies will not be ready to resume expansion of their systems until usage grows into the 80 percent range or significant new applications are deployed.

    The third and most critical driver to the current infrastructure slowdown--which is hitting networking, client and server purchases alike--is the budget lockdown and project deferral that many companies put in place during the first half of 2001. Starting early this year, most user companies put all new IT projects on hold. Our research shows that most new projects are likely to remain on hold through the end of 2001. As a result, IT departments have drastically cut purchases of new servers.

    We expect the excess capacity and secondary market glut to have been "digested" and the demand for nondiscretionary infrastructure enhancements to rise by the end of 2001. More importantly, we expect the demand for infrastructure to support new projects will also begin to surge again.

    How the turnaround will begin
    Although many projects have been put on hold, key CRM (customer relationship management), supply chain and even e-business projects have not often been totally eliminated. Instead, IT and business users are being forced to more clearly evaluate the cost/benefit trade-off and focus attention on projects that have the highest potential for near-term payback and/or significant competitive advantage in the market. In fact, we expect traditional brick-and-mortar companies to lead the charge (or at least the slow jog) back to e-business. However, rather than blindly allocate funds to "anything e," these investments will be highly targeted.

    We expect the first indications of a market turnaround will be the release of targeted projects currently on hold--including selected Windows 2000 migrations and targeted optimizing of CRM and commerce chain management arrangements by the end of 2001 or early 2002. The Web market will take longer to recover because of the cynicism now directed at e-business and the confusion over the best strategies to use on the Web.

    See news story:
    PC sales flu infects servers
    In the near term, the slowdown will have interesting effects on the adoption of new technologies. As vendors attempt to gain competitive advantage in a stagnant or shrinking market, we expect them to focus more attention on higher-margin, leading-edge systems in an attempt to gain market differentiation. However, many users may choose to stay with older technology--buying lower on the price curve to save money. This will likely force Intel, IBM and Compaq Computer to reduce the price of newer technologies (such as Itanium-based systems) more quickly than usual to drive demand.

    When the recovery does occur, we expect IBM, Compaq, Hewlett-Packard and, to a lesser extent Dell Computer--the Itanium vendor group--to be best positioned for recovery. Just as new project investment begins to flow a bit more freely, Itanium will begin to mature, providing a more unified "Wintelnix" platform. At least theoretically, users will have the ability to focus on a common hardware platform to run either Windows or one of a variety of Unix flavors.

    Therefore, Wintelnix may further isolate Sun's Solaris/SPARC platform. This, and the fact that Sun benefited heavily from the dot-com and telco boom markets that will be slow to recover, means that Sun will likely be slower to rebound.

    For users, this situation could hardly be more ideal. The current squeeze in server demand will force server vendors to offer excellent deals, but it will not be enough to drive anyone out of business. It will be a strong buyer's market this year and likely through the first quarter of 2002.

    However, many organizations are facing a soft business climate in their own industries, and spending must be done carefully. Users should review the projects they have on hold and pick the ones with the clearest and most immediate business advantage and cost justification. They should also plan carefully for exactly how they can implement these selected projects to gain competitive advantage in the marketplace.

    By focusing "rifle shot" doses of spending in these areas, some companies will be able to turn a tough economy into an opportunity and be best positioned to gain share when the economy eventually turns around. Leading-edge companies will start implementing projects earlier to get ahead of their competition and to exploit the present marketplace by getting the best deals on equipment.

    Meta Group analysts Dale Kutnick, Val Sribar and David Cearley contributed to this article.

    Visit Metagroup.com for more analysis of key IT and e-business issues.

    Entire contents, Copyright ? 2001 Meta Group, Inc. All rights reserved.