Parting Ways (Redux)

With rumors  of the Rock project (intended to be the next generation SPARC microprocessor) being cancelled two years after Jonathan Schwartz’s announcement on his blog,  Sun Microsystems may have finally conceded its independence as a systems vendor before its impending acquisition by  Oracle (see Parting Ways).

Perhaps the should not come as a surpise with the recent  launch of  mid-range systems from IBM, Dell, HP, and Sun based on the Intel Xeon 5500 quad-core microprocessor  (code named Nehalem) with claims of being half the cost and 1.7 times the performance of a comparable SPARC-based system. Likewise, at the high end, where Sun is the preferred platform for  Oracle databases and applications with its current Niagra-based servers, there are Oracle benchmarks that show that blade servers based on the the previous generation Xeon processors scale well  to “Big Iron” mainframe performance.  Even IBM has had to re-evaluate its commitment to its Power microprocessor architecute (as has HP with its commitment to the Itanium microprocessor).

With no apparent edge on the systems side of the business, what will Oracle do once Sun shareholders vote on July 16?  Oracle estimates that Java and Solaris are the crown jewels that will be quickly exploited due to tighter integration with the company’s existing database and enterprise middleware offerings. Such retooling could provide Oracle a significant competitive advantage over such rivals as IBM (see article) and Microsoft (see article), as well as contribute $1.5 billion to Oracle’s operating profit–excluding charges and other items–in the first year after the acquisition, increasing to over $2 billion in the second year.

Moreover, Larry Ellison said the acquisition gives Oracle the ability to  engineer an integrated system – applications to disk – where all the pieces fit and work together with customers benefitting as their systems integration costs go down while system performance, reliability and security go up (see article). However, Oracle already has such an arrangement with HP to build a grid of Oracle database servers with HP Exabyte servers.  Current industry speculation suggests that HP may buy Sun’s hardware, storage and integrated cicuits intellectual property outright after the Oracle acquisition is final (see article).

This still gives Oracle an incentive to continue as the steward for open specifications focusing on the Java Community Process and OpenSolaris. On the other hand, all may not be as rosy for colleagues working on open source initiatives (such as the NetBeans developer tools and the Glassfish  application server) that compete with more popular offerings such as Eclipse and JBOSS.  Such projects could soon be on the chopping block on top of the 6000 job cuts already announced by Sun (see article).

Besides the obvious layoffs involving  the systems side of the business, the software side will also come under close scrutiny with salaries that exceed those paid by Oracle.  Oracle has also earned a reputation for a military-style discipline that may come as a shock to more relaxed Sun employees.

In the end, Sun Microsystems became a victim of its own success. Like other successful startups like Silicon Graphics and Netscape Communications, the company differentiated itself with innovations such as Java and networked computing, but the company could not adapt as the times changed (like the dot com bust) and the technologies it evangelized became commodities that could be easily duplicated or replaced with ease and lower cost.