IBM/HDS Announce Storage Alliance/Disk Drive Venture

By Charles King

IBM and Hitachi Data Systems (HDS) have announced plans to form a multi-year strategic alliance to accelerate delivery of the companies’ advanced storage products and technologies to market. The companies plan to jointly research and develop next-generation storage networks, systems, and solutions based on the Common Information Management (CIM) standard advocated by the Storage Networking Industry Association (SNIA), and will also pursue a common approach to storage virtualization based on IBM technology. While the companies will work together on a variety of initiatives, they will continue to compete for market share in the enterprise storage space. IBM and HDS also announced a non-binding agreement to form a standalone joint venture that will integrate all of both companies’ various hard disk drive (HDD) assets including research, development, manufacturing, sales, and marketing. Upon completion of the deal, HDS is expected to hold 70% of the venture and to pay IBM for its HDD assets. No estimated financial or timetable details were included in the announcement.

While the IBM/HDS agreements may appear dramatic at first glance, they can also be seen as logical steps in the continuing evolution of data storage. While storage was once considered largely immune to the economic forces that were driving the rest of the tech sector into the ground, that notion went down in flames as storage players issued a stream of dim or dismal earnings reports. Among major players, fierce competition and aggressive pricing have led IBM and HDS to notable gains in market share over the past year, but both companies still trail storage solution leaders EMC and Compaq. EMC’s Wide Sky initiative, which the company is using to drive heterogeneous storage management standards through sharing its APIs, received a boost when Compaq signed on for the ride, leaving IBM and HDS on the outside looking in at a party neither particularly wanted to attend. What to do? Why not join hands with SNIA to drive CIM and carry the industry standard flag? This strategy allows IBM and HDS to claim a bit of moral high ground at they same time they pool resources to increase their individual competitive positions.

What about the IBM/HDS joint venture? In 1956 IBM introduced the industry’s first magnetic hard disk memory drives in its RAMAC computers, and the company has been at the forefront of storage innovation since then. But making a buck on disk drives has been tough in the extreme, since performance leaps have pushed drive technologies toward commoditization even more rapidly than microprocessors, increasing the economic pressure on everyone involved. By offloading its disk drive group (along with 15,000 or so employees worldwide) to an HDS-led joint venture, IBM can gracefully decrease its exposure in a risky sector, while keeping a finger on the tiller at the same time. So what does HDS get out of the deal? A literal host of IBM talent, products, and patents that could easily vault the company into a disk drive leadership position; not a bad place to be as powerful data storage solutions continue to be widely deployed. In fact, Hitachi may have visions of modeling itself after Intel in providing OEM commodity products for hard drive-hungry manufacturers of every kind. Overall, we see this venture translating into probable good news for consumers and businesses over the long term, but it is likely to be an additional thorn in the side of storage solution vendors.

Sun Creates Security Czar

By Jim Balderston

Sun Microsystems announced this week that it had appointed Whitfield Diffie as Chief Security Officer and also appointed Joanne Masters Director of Sun’s Global Security Program Office. Diffie joined Sun in 1991, and is most widely know as a cryptography and security specialist. He invented the Public Key Encryption schema and has been at the forefront of encryption and security-related discussions for at least two decades. Sun said the objectives of the new appointments are to advocate for Sun’s security offerings, educate partners and customers on security issues, improve ties between Sun and existing security organizations, and address future customer security needs.

Sun has always touted its products as being more secure than those of its rivals — most notably Microsoft — so in some ways this announcement comes as a bit of surprise. When one considers Sun CEO Scott McNeely’s mocking of a recent Microsoft security initiative, it makes this announcement all the more curious, at least from a public relations point of view. But regardless of the appointments’ PR value, we see this as the latest instance of an increasing trend within the IT vendor community toward integrating security at much lower levels in products; from the ground up, as it were. Companies like IBM have been laboring mightily to build security features and frameworks into their wide array of offerings, providing a more integrated security focus than many other vendors. However, to date, most IT vendors are still working under the paradigm that security is best left to companies specializing in this arcane and complex area. This, in turn, has encouraged and promoted the concept of security as a perimeter defense, a model that is rapidly being shown as inadequate in increasingly interconnected enterprise networks, where the more complex a network is, the greater its number of possible weak points.

As a result, we are seeing (only slowly) vendors’ increased interest in integrated security. Microsoft is a glaring example of such vendor behavior. The company’s utter and complete disinterest in security issues historically has always been a point of mystery to us, especially now that each and every security vulnerability is broadcast worldwide within hours of its discovery, thanks in large part to the Internet. Microsoft has apparently issued internal directives from the highest levels indicating that security is now a top concern. Sun has taken the next step and appointed a widely respected security expert to head its own increased focus on security. Going forward, the question that must be addressed by Sun and other vendors is how to provide its customers granular security solutions for an increasingly uncontrolled “melting pot” environment of near-universal connectedness.

The Goose or the Eggs? IBM Claims Patent to ebXML

By Jim Balderston

IBM sent an email to the Organization for the Advancement of Structured Information Standards (OASIS) a few weeks ago that came to light this week. In the email, IBM claimed to have one patent and one patent application that directly relate to ebXML; specifically the compliance of implementations of OASIS ebXML Collaboration Protocol Profiles (CPPs) and the Collaboration Protocol Agreements (CPAs). IBM claimed that Patent 6,148,290 covered these areas, and the patent application may also cover the CPPs and CPAs. These protocols allow for more complex messaging within the XML schema, and are broadly considered key to its implementation. In the email, IBM stated that it was willing to offer non-exclusive licenses under reasonable and non-discriminatory (RAND) terms and conditions. The email closed with information on how parties wishing to request a patent license could contact IBM. News reports of the email included a number of comments from companies working on or with ebXML standards; all noting that IBM had never indicated that it might have patents that would require royalty payments to IBM for implementations using the ebXML standard. It remains unclear whether IBM is going to enforce the patent and demand royalties, or if it will allow for less restrictive, non-royalty use.

To us, the choice for IBM seems very clear. If IBM wants to gum up years of XML development, it will insist on royalty payments for the use of the ebXML standard. It will kill the goose, as it were. IBM may in fact be in possession of patents that could require royalties paid for ebXML implementations. If so, and it chooses to require royalties, a significant number of companies will have to begin siphoning off revenue to pay IBM’s tax. Or, they can stop developing in ebXML and move in a yet-to-be-defined direction, thereby limiting ebXML implementation to IBM and those willing to pay the freight. Perhaps in some quarters of Big Blue, collecting fees and “owning” ebXML seems like a fine idea, indeed. It is a powerful and compelling technology, one that will play a crucial role in the realization of ubiquitous access to information that many industries have been promised yet have not realized to date with their IT investments. Perhaps IBM believes that owning this technology will give it a leg up on all other vendors, and press the case that if a buyer doesn’t buy from IBM, it isn’t getting the best technology. But as IBM’s countless partnership agreements show, IT buyers have habits, brand loyalty, relationships, and a host of other non-technological reasons to go with non-IBM vendors. Some people in IBM — those forging these partnerships — understand the market has many different reasons for buying non-IBM products, and if offering such as part of larger IBM sales is what it takes to get the order, than so be it. For simple, IBM-centric financial reasons, IBM should resist the urge to hoard ebXML. On a broader field of view, IBM must also recognize that the universal adoption of ebXML is not a lock at this date. While a great deal of enthusiasm surrounds XML and its implementations, it could be derailed if IBM moves forward with a harsh licensing scheme. In such a scenario, ebXML development would slow, non-IBM innovations would wither, and the pace of ebXML adoption would be severely hampered. If history is any example, it is more likely that IBM understands that unprotected patents aren’t worth the lawyers’ fees it took to file them, while exercising patents can ensure stewardship rights in the way their technology is broadly adopted. Any possible return in licensing revenue could kill the XML goose. A fine first meal, perhaps, but a paucity of golden eggs down the road, to be sure.

HP/Compaq Merger Approved According to Preliminary Tally

By Charles King

Hewlett-Packard announced that a preliminary tally of shareholder votes from the March 19 special election has approved the company’s proposed merger with Compaq Computer. The preliminary tally prepared by independent inspectors of election showed a margin of approval of roughly 45 million votes, or nearly 3% of the total. An insignificant number of votes remain to be counted. Votes for the merger totaled approximately 837.9 million. Votes against the merger totaled approximately 792.6 million, almost half of which were affiliated with the Hewlett and Packard family members and foundations. Shareholders not affiliated with HP’s founders’ families and foundations voted for the merger by a margin of roughly 2:1. If merger opponent Walter Hewlett demands a recount of the votes, both parties will participate in reviewing the proxies, a process that HP expects could take seven to ten days to complete.

The last two weeks’ worth of headlines concerning the HP/Compaq merger contain enough rumors, innuendoes, private voicemails, judicial proceedings, claims, counterclaims, and generic finger pointing to fuel at least three episodes of “Days of Our Lives.” Beyond the ongoing soap opera that this ambitious business process has become (largely due to the curious public and private machinations of HP scion Walter Hewlett), some light has begun to penetrate the gloom with the preliminary tally of shareholder votes. While some have pointed out how slender HP’s approximate 3% margin of victory appears, we suggest a different measure be applied. At this point, Walter Hewlett’s attempts to derail the merger depend on convincing a judge that HP illegally persuaded Deutsche Bank to vote its 1.7% of company stock in favor of the merger. Even if those votes are thrown out, HP should still have more than enough votes to succeed.

Does this mean Walter Hewlett will own up to the loss and withdraw gracefully? If the past is any measure, we believe that is unlikely. Despite the unthinkable damage he may do the company he professes to love, Mr. Hewlett is likely to battle on, imagining himself to be a valiant Don Quixote battling against infidel hordes, instead of a simple fanatic who, to paraphrase Santayana, redoubles his efforts when he has forgotten his aims. If the preliminary vote is finalized, where does this leave HP? On April 9, the company released a preliminary roadmap for life after the merger that offered an intriguing view of how the company would leverage Compaq’s considerable intellectual property and management expertise. What will be most important, in our view, is for HP to move forward intelligently, humanely, and decisively in its reorganization. Such an effort should allay the concerns of HP’s customers and address the fears of anxious employees. Despite the difficulty, unpredictability, and sometimes sheer lunacy the past few months have served up, now the real work begins.

The Sageza Group, Inc.

836 W El Camino Real
Mountain View, CA 94040-2512
650·390·0700 fax 650·649·2302
London +44 (0) 20·7900·2819

Munich +49 (0) 89·4201·7144
Amsterdam +31 (0) 35·588·1546

sageza.com
info@sageza.com