Market Roundup

March 7, 2003




EMC Inks API Deals with Veritas and HDS

IBM Announces $1 Billion On Demand Win

Intel Announces “Unwire” Centrino Campaign, T-Mobile Agreement

HP Announces OpenView Support for Web Services




EMC Inks API Deals with Veritas and HDS

By Charles King

EMC and Veritas this week announced an agreement to exchange a wide range of application programming interfaces (APIs) for storage system and storage software technologies. The agreement extends access to APIs the companies previously exchanged, and includes new APIs relating to storage arrays, volume and file management, and replication features. In addition, the companies will define cooperative support levels so that end users of the respective storage system and storage management products can receive joint support for the integrated product offerings. In an unrelated announcement, EMC and Hitachi, Ltd. announced a settlement of all pending patent infringement litigation between the two companies, as well as five-year agreements which include patent cross-licenses and mutual releases between EMC and Hitachi, Ltd. and its subsidiaries, Hitachi Data Systems Corporation (HDS) and Hitachi Computer Products (America) Inc. Along with this settlement, EMC and Hitachi also agreed to a framework for exchanging technology in the form of storage-related APIs, which is expected to be finalized in coming weeks. Under terms of the agreements, Hitachi agreed to make balancing payments to EMC. The remaining terms of the cross-license agreement are confidential.

From a practical standpoint, the API deals EMC arranged with Veritas and HDS have the potential to provide the three companies’ customers similar benefits. While a good deal of ink and heat (if not much light) have been expended on upcoming Common Information Model (CIM; aka Bluefin) standards that will reportedly ease the management of heterogeneous storage environments, the fact is that CIM standards are likely to arrive piecemeal and that fully functional CIM-compliant solutions are many months, if not years, away. Additionally, it is unclear just how much backward CIM-compatibility storage vendors will build into their solutions. If CIM standards are simply treated as a nifty feature for driving new storage product sales, their role in easing systemic storage management pain will be diminished. Until these issues are clarified, storage management can be notably improved by vendors actively pursuing and leveraging API swaps. How significant are these two agreements? What is most interesting about the EMC/Veritas deal is that it represents the first such agreement between EMC and an ISV; all other EMC API swaps have been with hardware vendors such as Compaq and HP. EMC’s long-standing role in enterprise storage and Veritas’s position as a major storage management player suggests that the deal has the potential to be highly beneficial for both companies and their myriad customers.

The EMC/HDS agreement is a bit harder to parse out and requires some reading between the lines, since the API language is buried in the larger commentary on the companies’ legal settlement. EMC and HDS have been fierce competitors for years, and HDS has been responsible for significant erosion of EMC’s enterprise storage market share. On the legal side of the announcement, HDS’s agreement to make balancing payments to EMC (with no similar agreement by EMC) is the most significant detail, suggesting that HDS likely decided that suing for peace would be less damaging than pursuing ultimately unsuccessful litigation. Given this, the companies’ API agreement could reflect a simple sweetening of the settlement pot or an attempt by HDS to interject a bit of good will into the proceedings. How much good will is involved will determine just how positive this agreement will be for end users. Overall, an agreement that follows the letter rather than the spirit of cooperation will be largely useless to enterprise storage customers.


IBM Announces $1 Billion On Demand Win

By Jim Balderston

IBM announced this week that it has signed a $1 billion, six-year deal with AXA Group to provide AXA with On Demand computing capabilities using IBM technology and products. AXA is a worldwide financial protection and wealth management company. Under the terms of the deal, AXA will retain all management control of its IT infrastructure, while IBM will help AXA consolidate its server, mainframe, and storage capacity. The announcement is the largest IBM has made since introducing its On Demand initiative last fall.

It is an old saw that the best way to be successful in business is to find a parade and get out in front of it. Such wisdom hints at a basic business idea: finding where your customers are going and putting yourself in a position to help them get there. The advantages of such a strategy are obvious: less resistance from clients that are already on the path you propose to implement, synergies in existing and new technology deployments, and a common vision of the end result (and success) of deploying such technology solutions. IBM’s On Demand initiative essentially offers enterprises the opportunity to consolidate existing servers, storage, and applications and to provide additional support with IBM Capacity On Demand (COD) and hosted services as needed to give a fluid, dynamic computing capability across the entire enterprise. Leverage existing assets and tie them together with a snazzy new ribbon, thereby offering your clients a comfortable ride down a road they may or may not be aware they are already on, is a much easier sell than the old rip and replace offerings.

On Demand is by no means fully defined. The key to IBM maintaining ongoing traction in the marketplace will largely rely on the company’s ability to continue to extend and improve on the concept. By offering enterprises the option of maintaining complete control over its infrastructure — like the AXA deal — along with offering an outside-the-firewall option in which IBM provides additional computing assets as needed in real time, IBM steps lightly around the old outsourcing bugaboo. While many IT departments simply do not cotton to the idea that they cannot go down into the server room and see all of their assets blinking and humming in air-conditioned comfort, others are becoming increasingly comfortable with the idea that someone else can manage at least parts of their infrastructure safely and securely. By offering a flexible set of options, IBM keeps its markets as broadly defined as possible, and allows itself to gain entry into enterprises that may initially want to maintain complete control, but over time may find that letting some capacity to reside offsite is more and more preferable. All in all, keeping On Demand flexible, offering a wide range of deployment and purchasing options to suit myriad members of its customer base, is an approach that we believe will play out to IBM’s favor in the coming years as these services become an increasing part of mainstream business computing.


Intel Announces “Unwire” Centrino Campaign, T-Mobile Agreement

By Charles King

Intel has begun a worldwide advertising campaign to support the March 12 launch of the company’s Centrino mobile technologies. Centrino includes a new Intel Pentium-M mobile processor, related chipsets, and integrated 802.11 wireless network capabilities designed to support wireless notebook PCs. In a separate announcement, Intel and T-Mobile USA announced a joint marketing campaign to promote T-Mobile’s WiFi (802.11) high speed wireless Internet service. As part of the campaign, T-Mobile announced new monthly pricing plans, including an annual agreement that provides users unlimited T-Mobile HotSpot service for $29.95 per month. HotSpot is currently available at 2,100 U.S. locations including Starbucks coffee houses, national and regional airports, and American Airlines Admirals Clubs. The service will also be available in new locations as they are deployed, including additional airports, Borders Books and Music stores, and United and Delta Air Lines clubs and lounges. During a related press event at San Francisco International Airport, Intel Chairman Andy Grove said the new Centrino chips were “second only to the introduction of the Pentium” in terms of importance to the company, and “will help jump start” a shift toward wireless computing.

Intel is putting a great deal of energy into the Centrino promotion, and though the actual significance of the resulting sound and fury may be uncertain at this point, we are intrigued by some “unwired” implications. Could Centrino really be as important to Intel as Pentium? Will it actually inspire a shift toward wireless computing? A bit of historical and current market reflection might be in order here. High tech industry graybeards will remember a time when PCs were anathema to enterprise IT managers, with the undeserved rap that they were mere playthings that could clutter or even cripple company networks. But a funny thing happened on the way to paradise, with workers smuggling PCs into the workplace to ease and enhance their jobs. Bottom line: end users were way ahead of experts in recognizing the value of personal computing. Shift ahead 20+ years and one can see a similar evolution taking place. Over the past three years, the growing performance and capabilities of notebook computers have driven them further and further into the business and consumer markets, nearly reaching parity with desktop PC sales. At the same time, roll-your-own wireless networking has become hugely popular both at work and home. In other words, much of the market is both well tilled and fertilized for integrated mobile/wireless products.

That being the case, is Centrino assured success? Intel and its PC vendor partners certainly hope so. The high-end graphics/video capabilities that were recently supposed to drive consumers to upgrade PCs have largely (and deservedly) nosedived earthward, and business desktop sales (except for notebooks) have been softer than a feather duvet for more than a year. Centrino’s leverage of two popular computing trends should give it supple enough legs to travel blithely between both wireless-enabled homes and businesses. So what about the Intel/T-Mobile connection? Given its design, we believe the T-Mobile service is likely to find initial success primarily among business travelers and power users, but could be buoyed as HotSpots are deployed in more consumer-friendly locations. At the same time, the significant performance capabilities and savings in wiring costs promised by WiFi networks inspires us to imagine a heady time when these solutions will be widely deployed in schools, libraries, shopping malls and myriad other public locations. By easing the rigors of the Last Mile, Intel and T-Mobile are likely to assure that more travelers will take the trip.


HP Announces OpenView Support for Web Services

By Myles Suer

Carly Fiorina announced at BEA World that HP will provide technology to help address Web services management issues. While asserting that Web services overcomes IT complexity, unifying systems and accelerating a return on IT investments, Ms. Fiorina stated that HP’s OpenView technology will eliminate issues that might otherwise inhibit the growth of Web services by allowing customers to provision and manage Web services via a consistent management interface for  both J2EE and .NET components. HP introduced the OpenView Web Services Management Engine, which allows customers to intercept Web services requests and actively manage the service apart from the platform where it resides. HP also announced OpenView Transaction Analyzer, which uses APIs, co-developed with BEA to monitor Web services applications.

As with virtualization technology, management applications will play key roles in determining and delivering effective Web Services deployments. HP, IBM, and BMC have all positioned themselves to assist customers in managing Web Services connectivity, though not for especially altruistic reasons. While using Web service technologies to leverage enterprise applications for additional purposes offers some compelling value, the possibility also exists for these new systems to compromise existing services and business processes by overuse. HP appears to be developing preventative solutions for this problem before it presents itself, an intelligent move in our opinion. We also think that it is wise for HP — no longer being in the application server business — to leave all doors open, including one for .NET. This allows HP to play with four (and possibly more) application server vendors (BEA, Oracle, Microsoft, and Sybase) and should provide the company myriad opportunities to offer its management services as effective means for dealing with the most heterogeneous enterprise computing environments.

OK. So far sounds great. HP is positioning itself to play manager for most Web Application Servers. The question the industry is really begging to understand is: Is this enough? HP shocked the industry by shutting down their own Web Application Server business (the Bluestone acquisition) last year and positioning themselves as the Switzerland of systems developers by announcing support for just about everyone else’s J2EE and .NET servers. This strikes us as “low risk” (not usually something we find exciting), but it may represent a stroke of genius. It is our belief that Web Application Servers are on the same commoditization path that has been trodden by Web and email servers. The result of complete commoditization has been a continuing devaluation of these solutions, which has effectively destroyed companies that relied too much on product revenue alone and not enough on the services and tools sales that grew around them. Overall, we see HP’s low risk approach as a potentially big gamble that will test both their technological acumen and their crystal ball.



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


Copyright © 2003 The Sageza Group, Inc.

May not be duplicated or retransmitted without written permission