There's something Kafka-esque about covering the IT business these days. With all the major changes lately, tracking the industry is bit like waking up with a new face every morning. Industry consolidation has reached such a fever pitch that "here today, gone tomorrow" could justifiably be called the new motto of the industry, and the web of vendor interrelationships has assumed the shifting contours of sand dunes in a stiff wind.
The latest jarring event is the elevation of the Compaq/Microsoft/Intel triple entente to superpower status, which became official with Compaq's merger with Digital Equipment Corp. These three companies are closely interrelated in their strategy for bringing low-cost, high-volume clustered NT servers into the data center en masse.
Intel and Microsoft must be overjoyed. There's now a vendor in place in time for the release of Merced with all the engineering, financial, and technology resources to make mission-critical, multiprocessing applications run on a unifying Intel architecture under NT. The pressure on competitors will be immense when Merced begins to wipe out Sun Microsystems' and Silicon Graphics Inc.'s pure Unix hardware businesses sometime by 2001. (The launch of the infamous "Intel Inside" advertising campaign some years ago made it clear that Intel was no longer content being a silent partner in the information revolution. With Merced, it will arrive at the high end of computing in force.)
On paper, Compaq is cut out for the job. Having made inroads into small and midsized corporate accounts with its well-received ProLiant SMP server line, Compaq can now compete not only in the corporate computing stratosphere with high-end Tandem and Digital technologies but also offer the services to back them up. The result is that IBM and Hewlett-Packard have finally shaken a waning Digital, only to find it replaced by a more aggressive, wealthy, and complete adversary with a significantly better understanding of marketing. (Compared to Digital, any understanding of marketing is an improvement.) Furthermore, Compaq may now hold an edge in the market for Windows NT consulting services--according to analysts, the most prized jewel in the company's crown--which is sure to expand rapidly as enterprises accelerate their deployment of NT beyond the department without expanding the number of people required to support it. Such services will be in high demand considering that unlike MVS, AS/400, or even VAX/VMS, NT is hardly bulletproof.
The reality, however, is that the arms of the Compaq/Microsoft/Intel pincer movement are not operating entirely in sync. Now that Compaq has entered the high end with proven Tandem transaction-processing technology and expertise--not to mention an excellent opportunity to be first to market with Merced-powered clustered servers--where does that leave Microsoft? Surely the company that hired Jim Gray, is virtually rebuilding SQL Server from scratch to enable terabyte-class support, has released an "enterprise" edition of NT, and is on the verge of wiping out the OLAP industry as we know it with the Plato OLAP server won't be willing to relinquish its claim on the data center. (Example: Oracle's and IBM's parallel architectures for NT were absent at Microsoft Scalability Day in May 1997. Only Tandem--no threat at the time--participated.) If that's the case, despite today's wine and roses, Microsoft and Compaq may be headed for a showdown.
This issue of Database P&D covers this new industry dynamic from various angles. On the technology side, Gregory Pfister's analysis of the trade-offs among parallel-processing architectures ("Sizing Up Parallel Architectures,") reveals why clusters are the official architectural model for NT scalability--system failures, which NT can offer in spades, are contained on a single node. Furthermore, in "The Viper Strikes" , Tim Matthews describes the nature and meaning of Microsoft Transaction Server, the middleware mainstay of Microsoft's component-based strategy. On the business side, columnist Alan Simon describes his experiences at Digital in the late 1980s and early 1990s and explains why the erstwhile powerhouse went astray. Lastly, in an exclusive interview ("Compaq's New World,"), Compaq vice president of database and business applications David Rodgers discusses his company's invigorated enterprise strategy and the new direction of the database industry.
As a new editor-in-chief, I want to ensure that Database P&D is addressing the issues that concern you most. In lieu of a formal editorial survey, I welcome any comments on the current focus--or lack of focus--of the magazine. You can reach me at jkestelyn@mfi.com.