February 13, 2004
Sun Microsystems held its Analyst Summit 2004 and NC Q1 04 product briefing this week in San Francisco. Amongst the many announcements was the new Sun Fire V20z server based upon the AMD Opteron processor and the new Sun Fire B200x dual processor x86 blade. As part of the Sun and AMD alliance announced in November 2003, both companies are collaborating to optimize the performance of Solaris, Java, and development tools for Opteron-based systems. Other announcements included the Sun N1 Grid Provisioning Server 3.1 Blades Edition, which allows customers to seamlessly mix, match, and manage x86 and UltraSPARC blades running either Solaris or Linux as one pool of resources. The company indicated the latest edition is up to twice as fast as the previous and can reduce server deployment time while simplifying blade server management and increasing server utilization. In addition, the new offering expands support for x86 blades and the Sun Fire B10n Content Load Balancing Blade. The Sun Fire V20z features dual AMD Opteron processors and supports Solaris x86, Red Hat, and SUSE Linux with prices starting at $2,795. Separately, Sun announced the Java Enterprise Developer Promotion for the Sun Developer Network (SDN) community, which is available to SDN members in the U.S., as a subscription based service for $1,499 per year for three years. The promotion includes, among others, the Sun Java Studio Enterprise, a development platform for the Java Enterprise System; Sun Studio, an IDE for C, C++, and Fortran applications; Solaris 9 for x86; a Sun Fire V20z Server; Support Services, including next day, on-site support as well as online forums and technical support, and 90-day installation support and software updates. The promotion runs through June 30, 2004.
I’ll take Great Predictions in Computing for $200, Alex. The answer is “32-bit computing is dead.” The question is “How did Sun misread the tea leaves?” OK, much has been written about one of the most interesting, entertaining, and nonetheless innovative IT vendors to grace the marketplace. With the annual gathering of Sun’s closest friends, partners, and perhaps worst enemies (depending upon your point of view), it is easy to pick apart the meteoric rise and sudden fall of the Copernican company. However, this analyst summit combined with the announcements of the last few months provide a fresh perspective that is in sharp contrast to a company whose market approach was increasingly, “our way, or the highway.” While the Open Systems mantra is still evident, it is clear that the company has learned a few lessons, and taken corrective action, so that it can broaden its appeal to those in the marketplace who do not find themselves in the give-me SPARC/Solaris-or-give-me-death faction. Sun’s strategic embracing of Opteron has given rise to some interesting product offerings and a marketing message that is predicated on value-added systems-focused computing that encompasses x86 32-bit, UltraSPARC 64-bit, and x86 32- to 64-bit migratory computing. Choose Industry standard x86 architecture or UltraSPARC and then choose Solaris or Linux, and deploy your applications. This sounds a lot more “open” than the Open System solutions of recent times.
Clearly looking to avoid repeating a mistake, Sun is running with Opteron early in the game with the hope of establishing itself as the Tier 1 provider of Opteron-based systems. By adding its extensive networking expertise and operating environment, these solutions are well positioned to stand out from the inevitable market flood of white box servers that happen to have an Opteron inside them. But perhaps one of the most interesting facets of these new systems is Sun’s foray into a subscription-based model for delivery of the complete system. Rent the software for $1,499/year for three years, and get the box free. Although this may seem an abrupt departure from Sun’s historic position of driving the cost of software to zero, it does illustrate Sun’s embrace of a recurring revenue model, where the Sun server is akin to a “free” cell phone that is part of a multi-year service agreement. While it remains to be seen whether this approach will generate sufficient market interest to percolate through to other Sun offerings, it is an example of challenging the marketplace through technological as well as marketing acumen. For a company that has such a legacy in R&D and engineering, it is refreshing to see this energy being reapplied not only to its own homegrown products, but to the industry standard architecture that not all that long ago was effectively abandoned by Sun. What’s that light breaking through the clouds? It may be a glimpse of dawning of a new era in which we may again see a rising Sun.
Comcast Corporation has announced a $54 billion bid to buy the Walt Disney Company, the third largest announced communications merger bid in history, trailing the January 2000 AOL bid for Time Warner, which was announced at $165 billion at the time. Comcast is the nation’s largest cable operator, with 21 million subscribers in markets across the country. Comcast said it has five million subscribers to its high-speed Internet service. Comcast had revenues of $18.3 billion in 2003, nearly all of that coming from its cable operation; a small sliver came from cable-TV channels it owns. Disney had revenues of $27.1 billion in the past fiscal year, with large percentages coming from its TV and cable networks, film and TV studios, parks and resorts, and a smaller amount from merchandise. The buyout offer was unsolicited, and comes at a time when Disney management is fighting contentious internal battles over both the control and the future of the company. Disney’s Board of Directors has said they will evaluate the offer; published reports indicate that it is being viewed as a hostile bid. Any buyout would be subject to regulatory approval.
It is both amusing and instructive to note that the Comcast bid is approximately one-third of the much-ballyhooed AOL Time Warner deal, one which in hindsight has largely been dismissed as the epitome of Internet hubris and megalomania. In that deal, AOL brought Internet subscribers to the table, Time Warner brought everything else: publishing, theme parks, music, film archives, production facilities for both TV and film, and cable networks and subscribers. Time Warner even had professional sports teams to throw in the mix. In short, the Internet boom-fueled minnow swallowed the whale.
In the Comcast/Disney scenario, we see a much more balanced and intriguing dynamic at work. While this deal is far from consummated, and in fact may never get past the initial sparring of lawyers and investment bankers, much less regulators, it is interesting to consider what might actually come to fruition is the deal does come to be. Comcast, unlike AOL, brings 21 million homes to the deal, 21 million last-mile subscribers who are already well versed in Disney offerings as they sit in front of the Comcast-powered set-top boxes on top of their TV sets. Back in 2000, AOL had millions more subscribers, but they essentially represented not a communications network but a large mass of people who hit AOL’s servers via a host of different routes that AOL (and for the most part Time Warner) did not own or control. In the Comcast/Disney scenario, Comcast would own both the means of production of content and the means by which to deliver to consumers. In contrast to the AOL deal, which basically brought distant subscribers who theoretically might someday consume TW content, this buyout has much more synergistic prospects. Back in 2000, AOL Chairman Steve Case spoke of the synergy between old media and new, something that was largely based on the leap of faith that AOL subscribers would somehow be given access to Time Warner content, something that has happened only to the slightest degree. In this week’s proposed buyout, we see a much higher probability that such a scenario would play out to the benefit of the newly created Comcast media behemoth.
The Justice Department has released a preliminary decision indicating that it would oppose Oracle’s proposed buyout of PeopleSoft, a $9.4 billion hostile takeover bid that Oracle launched last June. The staff recommendation is not final; the assistant attorney general handling the case will release a final decision on March 2. The Justice Department ruling comes a week after Oracle upped its bid by a third, in an attempt to capture wavering PeopleSoft shareholders. When the original acquisition plan was announced Oracle indicated that it would not continue development of PeopleSoft products, and would only offer a minimum amount of support for installed products going forward. Instead, it said it would offer PeopleSoft customers migration tools to Oracle products, eventually doing away with PeopleSoft applications altogether. Oracle has indicated that will continue to pursue the buyout and may take the Justice Department to court in an effort to pave the way for approval for the buyout.
While Oracle may hope and try to pressure the Justice Department to reverse its decision, historically staff recommendations of this sort have been upheld by department higher-ups. Many observers believe that this preliminary ruling means the end of the deal despite Oracle’s claims otherwise, but regardless of the outcome this is not the last we have heard from Oracle on the acquisition front. The company has no choice but to expand its horizontal reach. Simply selling databases in an increasingly competitive market will not assure the company a rosy future. With Microsoft on the low end, and IBM pounding away above, Oracle has no choice but to spread its wings across the application spectrum.
That said, perhaps Oracle will learn a lesson from the PeopleSoft exercise, one that might improve the company’s chances of completing future buyouts. We suspect that one of the main reasons that the Justice Department balked at approving this deal was the fact that in no uncertain terms this was not a technology buy, but a customer buy. The fact that Oracle made it clear it had no intention of continuing the PeopleSoft product line, and instead would offer PeopleSoft customers a migration deal they could not refuse, looked to us to be more extortion-like than anything else. Given the fact that these customers would be forced to either go with Oracle or do an entire rip-and-replace with another vendor left few feeling particularly pleased about their IT infrastructure prospects, largely because they we being given a Hobson’s choice. Perhaps the next time, Oracle might want to consider more than simply what is good for Oracle, and avoid the kind of forced-march recruitment of new customers that in most cases winds up hurting the market at large, regardless of the short-term benefits to the acquiring company.
HP have announced a call centre in Bangalore, India that will deal with U.S. consumer post-sales support. In addition to customer service, the centre will also serve as a facility for research and development on support processes and technologies that will then be deployed across HP’s network of partners for service and support. The centre represents the first wholly-owned contact centre for post-sales support for HP.
HP are just another in a growing procession of companies that have announced offshore contact centres and customer service operations, particularly in India. This news is inevitably positioned as yet another brilliant idea by the firm for lowering costs and improving customer service, thus its appeal to shareholders. Naturally, it is also portrayed as a win for Indians and the Indian economy and the future of the whole developing world gets to ride on its coattails. What has been boiling under the surface, however is an apprehension on the part of the actual employees, especially in the U.S. and the UK, about the implications of this trend for domestic workers and the glimmerings of what promises to be an emotional debate about what should be done about offshore operations and whether they should be discouraged or perhaps even penalized.
We think that what is going on in this case is that moving employees offshore is the logical extrapolation of virtualisation to its extreme conclusion. Virtualisation has been the sexy high-tech fashion for the last couple of years. At the highest level, virtualisation is about using physically and logically distributed resources to create a system that appears to be directly before the user, presenting them what they need with no more overhead in response than if that system really were physically local to the user. The reasons for virtualising a system are generally to lower costs, reduce complexity, and simplify manageability issues. It can also be used to help scale a system logically beyond its current physical limitations; hence the evolution of servers and storage systems that present themselves to applications as a single server or storage device, when in fact the application may be running across multiple servers or accessing data from many different storage devices simultaneously. What we do at the product level, we are doing to processes and now people as well. While these topics are not new to business – outsourcing of processes such as human resources or even parts of IT such as email are common, and factories have been moved offshore for years – it means that a UK user of a corporation’s products and services can interact with an office in Bangalore that yields the same results as if they had called a local or regional office. Of course, virtualisation in IT is an imperfect thing, incorporating as much art as science at times. The same is true for working with human beings, except that unlike their high-tech counterparts, human beings react to their environment and respond to their redeployment. They are much more sensitive devices than the most particular electronic device. This changes the equation considerably. For each virtualisation of human resource, there are, in fact, humans that are displaced.