February 25, 2005
IBM has announced its new eServer X3 architecture, which will be applied to the companyís 64-bit Intel Xeon-based xSeries servers. The new architecture is the product of a three-year, $100 million development effort and will allow customers to run both 32-bit and 64-bit applications on the same computer. More notably, the new chipset features virtual memory with low enough latency to do away with the need for physical L4 cache on the chipset. IBM says the x366 server outperformed similarly priced models from competitors like HP. IBM also says the virtual cache and the elimination of the physical cache will allow for lower price point for both two-way and four-way computing environments. The x366 eServer will be available within three months and will be priced starting at $6,999.
With the market for Intel-based servers continuing to grow, offering the X3 architecture presents IBM with an opportunity not only to capture part of that market growth but also to lasso in some revenue associated with the growing pains of that market. To be specific, the X3 architecture provides IBM the ability to offer lower-cost alternatives to customers seeking ways to consolidate their existing Intel-based server architecture while at the same time giving the customer the ability to begin or accelerate the migration from 32-bit to 64-bit computing.††
IBM has largely positioned this offering as a large enterprise product and rightfully so, for the most part. Enterprises with huge investments in the Wintel platform are seeking ways to cut costs and manage complexity through consolidation. The X3 architecture and products based upon it will provide those opportunities at substantially lower costs per server and processor. But we suspect that in time the X3 architecture will find its way into the mid-tier enterprise market as well, for some of the very same value propositions. SMBs are looking for ways to bring down the cost of computing cycles along with consolidating their increasingly complex server environments. As we have noted before, in many cases the products designed for the enterprise do in fact have a place in the middle tier market. If the X3 is all that it is advertised to be, we believe it will follow this pattern as well. After, whatís good for the goose is good for the gander.
IBM announced the IBM eServer iSeries Initiative for Innovation this week, a program designed to provide additional support to ISVs interested in porting their applications to the iSeries. IBM estimates that over 2,500 more ISVs will have access to an average of $50,000 worth of support and resources, including the availability of IBM experts in Rochester; access to software and On Demand business resources in five Innovation Centers worldwide; and a roadmap for developers, loaners, education, and co-marketing agreements. In addition to investment, the iSeries group borrowed a page from the mainframe group and released the iSeries Charter that outlines IBMís ongoing commitment to the platform, the community, and the customers. The launch coincides with Partnerworld, IBMís annual worldwide meeting for partners, and one of IBMís biggest and most popular events. As part of this program, IBM is promoting tools innovation support, meaning that IBM will be supporting a broader portfolio of tools solutions kits. This benefits two groups of ISVs. For those who already have applications for iSeries, the vast majority of which are still in RPG or COBOL, it means they will have greater flexibility in moving those applications to a Java environment and even integrating them with Microsoftís .NET environment. For ISVs who are porting applications for the first time, it means that they can use the tools of their choice and not be forced down any particular path.
Perhaps IBMís biggest problem for any one product is the success of all the others. Certainly this has been an ongoing problem in its server division. With so many good products to choose from, IBM must refocus from time to time, and this means that other platforms get less attention for a while. So it has gone for the iSeries. A perennial favorite and much-loved by customers, the iSeries has suffered a paucity of parental doting while IBM worked on getting other parts of the business such as UNIX, the mainframe, and storage out of their ruts and back to being top-performing units. Alas, while the product has continued to get better in that time, marketing has lagged behind. Although many within and outside IBM value the iSeries and appreciate its role, a new generation of customers has lost focus of what makes this platform so great thirty years after its introduction. The key to keeping a platform alive is the availability of customer applications for it, and ISVs are the lynchpin of the ecosystem. Without them, the technology would join the list of brilliant but underused and unappreciated systems that clutter the history of IT.
IBMís two-pronged approach of broadening ISV investment and restating its commitment to customers is the right first step in reorienting attention on the iSeries. Of course, there is always execution, and ISVs will want sales and marketing attention once they get those applications up and running, but the value proposition of the iSeries ó a computer that runs itself so you are free to run your business ó is an enduring value proposition, particularly for small and medium-sized businesses or remote locations of mid-sized and large ones.
This week Sun Microsystems, quietly and with little fanfare, indicated that it would start shipping the companyís v490 and v890 midrange servers with 1.35GHz UltraSPARC 4 processors. These systems had previously used 1.2GHz processors. Sun plans to equip the entire Sun Fire server line with faster processors and by the end of 2005 the products will use 1.8GHz UltraSPARC 4 Plus processor chips. Sunís higher-performance multicore processor chips will not be available until the 2006Ė2008 time frame. Besides chip upgrades on Sun manufactured severs, the company will start selling Fujitsu PRIMEPOWER server products, ranging from the single processor PRIMEPOWER 25O to the 128-processor PRIMEPOWER 2500. Sun is offering its own Sun Fire products with Solaris 10, but the Fujitsu PRIMEPOWER products will only be sold with Solaris 8 and 9. Sun and Fujitsu have had a twenty-year strategic relationship and in 2004 the companies agreed to jointly develop future server products, but these products are to use Fujitsu SPARC64 processors and technology rather than Sun UltraSPARC V. The companies plan to start selling co-developed servers under the band Advanced Product Line (APL) in 2006. In the meantime Sun will be approaching customers in certain geographies with the Fujitsu PRIMEPOWER offering.
Sun chose to incorporate the above messages into the company website, rather than convey the information through the more formal and visible mechanism of a public announcement or press release. We believe the company selected this approach because it is not that significant of an offering. Nevertheless this could be an important midlife product upgrade for Sunís UltraSPARC server customers. Sun may be attempting to fill a product performance void until the 2006 APL product availability, and remain competitive with customer upgrades, in the hopes of stemming Sun customer defections to competing platforms. By selling the Fujitsu PRIMEPOWER offering the stage is being set for the joint Sun and Fujitsu APL product line and what will very likely be a joint marketing effort. The midlife upgrade, sale of Fujitsu PRIMEPOWER, and joint Sun/Fujitsu effort might help keep SPARC alive, and allow Sun to commit greater resources to other opportunities such as the Sun Grid and open Solaris initiatives. Sun may also intend this midlife upgrade as the bridge to future UltraSPARC V, as the efforts evolve.
These actions may well help reassure Sun customers that the company will continue to provide competitive SPARC server platforms and a future customer growth and migration path, although this positive impact may be offset by other factors. For example, it appears that Sun multicore SPARC availability is running behind similar offerings from IBM and Intel. Customer viability and success of the joint Sun/Fujitsu APL product offering is still an unknown. Such issues leave the Sun customer base open to competitor FUD, which has already resulted in Sun customer defections. In the past there have been some vague rumors that Fujitsu might acquire Sun; rumors that were particularly well fueled by industry pundits. As the two companiesí SPARC product lines merge in 2006, such a marriage may become more likely. This may again become a Sun customer concern and a competitor opportunity. It will probably be 2007 before the realities and results of the Sun strategies are evident in the market and industry.
News reports indicate that the Cabir cell phone virus has made its way to the U.S. from the Philippines, according to security experts. To date, the fifteen variants of the virus have been found in largely benign forms, the most malicious carrying a payload that drains cell phone batteries. The Cabir infection is transmitted through Bluetooth connections to and from Bluetooth-enabled phones. The domestic infection was noted in a Santa Monica, California cell phone store display. Security officials said the infection was probably brought back to the U.S. by a cell phone user traveling overseas. In an unrelated event, news reports indicate that socialite Paris Hiltonís T-Mobile sidekick was hacked into, and its contents ó including stored phone numbers and photographs ó were posted on the Internet. A number of people whose phone numbers were stored on Hiltonís Sidekick have complained of receiving unsolicited phone calls. Finally, in a speech given at the RSA conference last week, futurist Paul Saffo noted that the information revolution is over, and that information is being replaced by media, which he said is now touching our lives in deeper and deeper ways.
In many ways, we would argue ó and have argued for a number of years ó that Saffoís observation is largely true. To us, the information revolution becomes realized when the technology and the habits it engenders become so much background noise in our lives. Just as the telephone and the television have become standard props of life throughout the world, we expect super-cell phones, animatronics, and ubiquitous computing to become as much a part of daily life as throwing laundry in the washing machine or shoveling snow from the walkway. As all of this whiz-bang technology becomes invisible, the revolution blooms in full flower.†
But not without costs. It was once noted that while the telephone conquers distance it takes as a price oneís privacy. The revolution has its negative impacts as well, as many of Hiltonís friends can now testify, with their private phone numbers disseminated worldwide in a matter of hours. The Cabir infection will no doubt be altered in the near future to contain a more malicious payload, and the increasing popularity of Bluetooth-enabled phones will speed its dissemination as well. Increasing interconnectivity comes with risks. In the corporate world, interconnectivity and transparency is supposedly everything. We wonder if the risks associated with this transparency are being fully explored and appreciated. Yes, security firms continue to offer new and different means to protect IT environments, but we believe they are not keeping pace with the demand for less and less friction between network participants. Moving forward, we believe that the solution to this quandary may well be the assignation of different values to information. Just as a dollar can have a variety of costs or values associated with it (a cash dollar spent differs from one spent on a credit card with 12% interest) so we suspect information will develop levels of value and cost in the coming future. Not all information is equal. Just ask Paris Hiltonís friends.
Lucio Stanca, the Minister of Innovation and Technologies for Italy, highlighted Lombardy and Milanís role as the motor of innovation in Italy this week and commented on a new e-government initiative for the region. Stanca is responsible for the modernization of Italyís information and communication technologies (ICT). In particular, he focuses on the development of the information society in Italy and its incumbent projects for government, business, and citizens. Stancaís appointment by the Italian Prime Minister, Silvio Berlusconi, as a government minister indicates the gravity with which Italy is approaching national digital development. Among other projects the ministry has in development, the region of Lombardy is hosting a project, Borsa Lavoro Lombardia (Work Exchange Lombardy). The site is Web-based and more, designed to help job applicants and companies find each other, and to provide information on training and skills development. Along with the Web site, the government is developing a call center, as well as information to be published on the terrestrial TV stations for those who donít yet have access to the Internet. The project is just one of many initiatives Stanca is driving for improving the e-infrastructure of the Italian government and driving Italy into the forefront of the digital age.
Prior to joining the government, Luca Stanca worked for over thirty years at IBM and at one time ran IBM Italia and later IBMís entire EMEA region. While a large IT company like IBM certainly presented challenges for an IT executive, modernizing Italian IT has been a different experience entirely. In a recent interview with an Italian magazine, Stanca described the Italian public sector as another planet in comparison to working for an IT company. Certainly he understands the possibilities of IT as well as the realities of customers constrained by budgets, old processes, and inherent organizational politics. Nevertheless, he has announced a recent spate of projects to achieve his goals. His aim to drive Italy forward digitally has recently focused on Milan. According to his office, Lombardy spends 600 euros per capita on ICT, which is about average for the EU, but is a bit behind North American spending of around 1000 euros per capita, and Stanca would like to see that grow. Milan is arguably the most-cabled city in Europe, with about 600,000 km of fibre optic cable, giving almost all of Milan access to high-speed lines. Only three of Milanís outlying communities donít yet have access to high-speed networks. Building and testing prototypical services in Milan is sensible, as this is the commercial center of Italy and full of inhabitants willing to experiment with the benefits of technology.
Most IT companies focus on the UK, Germany, and France when they approach European markets, as those represent the lionís share of IT spending. Italy is generally seen as a smaller market, harder for non-Italian speakers to penetrate, and more dependent on relationships for building business than most Americans are used to, which makes it a more daunting task for foreign companies. At the same time, since the dissolution of Olivetti Italy has not really had any major indigenous IT companies spring to the global level, and it has lost some of its advantage as a player in its own right, making it more dependent on global companies as IT suppliers. However, young Italians, like their other European counterparts, love iPods, Xboxes, DVDs and on-line chats. They are some of the biggest users of mobile technology and are taking to the Internet happily as Internet cafes and home computer penetration grows. There is certainly a generation of programmers and developers nascent in the mix. Stanca is banking on the interest of business people and high-tech aficionados to grow Italyís market for both Italian and external IT companies. What better place than Milan? It has been known as a design and fashion hub for years, and may now have a chance to become the high-tech hub for Italy if not for all of Southern Europe. Ci vedremo.