After Hewlett-Packard announced its intention to merge with rival Compaq Computer, HP heir and board member Walter Hewlett faced a crisis of confidence. Should he stand with the aligned forces of HP’s leadership or give council to fears of a future he could not envision? In his decision, Mr. Hewlett has labored publicly and vociferously to turn HP away from a deal he believes will be essentially fatal. In the latest of a series of full-page newspaper ads, Mr. Hewlett implored HP stockholders not to “trade HP’s crown jewel imaging and printing business for Compaq’s low-margin commodity computing business.” In a fundamental misstep, Mr. Hewlett’s opining that HP’s signature imaging and printing business will be compromised by Compaq’s essential focus on PC manufacturing misses or oversimplifies much of the synergy this merger generates.
A close examination of Compaq’s business shows that it matches or leads HP in PDAs, laptops, thin clients, low and high-end servers, data storage, and mainframe computing, not to mention the PCs Mr. Hewlett is so fixated on. HP, on the other hand, matches or exceeds Compaq in mid- and high-end servers, middleware and management software, networking technology, enterprise services and, of course, printing and imaging. These are areas where the combined company can achieve significant cost savings in R&D, marketing, and manufacturing with, if the deal is executed successfully, little risk of losing market position. Mr. Hewlett’s crisis seems inspired by a fear of the commodity nature of PCs. For a moment, let’s join him in ignoring the rest of the challenge and opportunity this merger represents to examine the nature of commoditization in the IT industry.
An important distinction to consider is the difference between commodity components and commodity products. Commodity components, such as cables, disk drives, memory chips, and now especially CPUs are defining and delivering architectures that reduce the cost of developing and delivering finished goods without compromising quality. Today’s computer vendors routinely use commodity components to stay competitive. An unprecedented event driving the HP/Compaq deal is the evolving commoditization of high-end CPUs being fueled by advances in Intel’s 32- and 64-bit architectures. Every major hardware vendor (with the exception of Sun Microsystems) plans significant product development based on these new Intel chips, and both HP and Compaq years ago announced plans to utilize Intel CPUs for all of their high-end computing needs. On the other hand, commodity products have little differentiation beyond brand and price because they fill simple, static needs. Personal Computers, clearly trending in this direction, are a concern for any PC vendor with an eye to having a future. Where Mr. Hewlett has gone wrong is in confusing commodity personal computers with the industry’s migration to commodity components. Far from continuing the trend of failing profit margins commoditization has brought to the PC marketplace, designing and building higher end computing products on industry standard commodity CPUs should help hardware vendors stabilize costs and eventually increase revenues across their enterprise product lines through value added hardware, software, content, and services. The reason for this is simple: The enterprise computing market is unlikely to stultify in a manner similar to the PC market because the needs of enterprise computing will continue to become increasingly complex in areas beyond the hardware’s basic architecture.
History is on our side here. Beginning in the 1970s and through the 1980s, vendors began integrating increasingly available commodity components into their largely custom-built products. However, industry leaders including Sperry, Burroughs, and Honeywell resisted the trend, clinging to low volumes and high margins, assuming they could bend the market to their wills. The results for these companies in the mainstream computing marketplace are self-evident. The cyclical nature of commoditization pressed on through the 1990s, and now consumer PCs have become the new “meat on the table.” Enterprise computing is a wholly different business, both literally and figuratively. We believe a merged HP/Compaq will be able to take greater advantage of the momentum growing behind Intel-based commodity CPUs than the companies could separately. Working together, the pair should be able to leverage their products and experience into an integrated solution set that meets enterprise computing needs and matches up well against the competition.
Does this mean it will be easy? Not at all. Every merger is complex and fraught with peril. The rancorous atmosphere inspired by Mr. Hewlett means the upcoming merger votes by HP and Compaq shareholders will be picked apart regardless of the outcome, possibly impeding the companies’ further efforts. In the end, this may suit Mr. Hewlett just fine, but we hope he learns something about commoditization, as he turns and faces the selfsame problem in his beloved printer business.
Hitachi Data Systems (HDS) announced HiCommand 2.0, a revision of the company’s data storage management framework. The new version of HiCommand features “link and launch” capabilities of Sun Microsystems’ StorEdge Resource Manager software, and adds Sun StorEdge T3 arrays to its portfolio of managed storage systems. Sun’s StorEdge Resource Manager software allows users to move across storage management software in multi-vendor environments to fully view storage resources including system identification, firmware revisions, controllers and physical disks, port identification, allocated and unallocated space and storage consumption. Designed as an open framework, HiCommand offers a Java-based GUI browser that can manage storage systems from any location. HiCommand provides a set of storage APIs that are being supported by ISVs including Sun, VERITAS, Microsoft, BMC, Computer Associates, and InterSAN. HiCommand can manage Hitachi Freedom Storage Lightning 9900 and Thunder 9200, the Sun StorEdge 9900 series and StorEdge T3 arrays. General availability for HiCommand 2.0 began in February 2002.
At a number of levels, this new version of HiCommand is hardly surprising. Resource discovery and management are key elements of storage solutions from virtually all enterprise vendors, and its addition to HiCommand allows HDS to be more competitive with the likes of EMC, Compaq, and IBM (Tivoli). Additionally, management solutions for heterogeneous storage environments are more a simple necessity than an option in today’s increasingly distributed and complex storage market. We should also point out that Sun’s StorEdge Resource Manager qualifies as a typical value-added response from a vendor to a favored OEM, in this case HDS, whose Freedom 9900 series are re-branded and sold as Sun’s StorEdge 900 line of SAN products. In nearly every case, HiCommand 2.0 simply reflects and supports the global cross licensing and distribution agreement Sun and HDS announced last August.
So is HiCommand 2.0 just a new rev of an old product, with a nod to one of HDS’s best buddies? We think there is a bit more here than initially meets the eye. For example, when Sun announced a series of new storage initiatives early in February, some industry pundits suggested that Sun’s improvements to its T3 RAID array line might endanger its HDS relationship, since HDS offers similar discreet products. If that were so, why is support for the T3 included in HiCommand 2.0? As we pointed out in our recent Sageza Competitive Review on IT alliances, properly formed partnerships allow vendors to extend their reach into areas and sectors that might previously have been beyond them. Simply put, a standard OEM agreement is tactical by nature in that it allows a vendor to quickly and easily fill a hole in its own product line. From our point of view, vendor alliances such as the one between HDS and Sun are more strategic and encompassing. Since HDS’s corporate reach is less extensive than Sun’s, the company stands to gain more now from supporting Sun’s RAID products than it does in promoting its own. At the same time, HiCommand’s support of T3 arrays provides Sun an integrated management platform to leverage for its own storage efforts. By working closely and strategically, HDS and Sun can extend and expand the size of their own market footprints. That is good business from any standpoint.
RealNetworks (RN) made two announcements this week focusing on the television marketplace and the traditional PC space. Intel, as an extension of their existing agreements with RN, will distribute the RealOne Player on Intel’s newest PC desktop boards. The inclusion of the RealOne Player will enable computer manufacturers and system integrators to deliver it as an integrated part of their PCs thus allowing consumers to have access to digital media right out of the box. RN also announced that it sees its future in television, not the computer, and offered some insight into its plans to focus it future direction on offering services in addition to software.
RN has been fighting a not so cold war with MS to get RN products onto MS dominated desktops. What has vexed RN of late though is that MS’s own technology has supplanted RN’s as the default (if not preferred) technology for streaming and digital media in the consumer marketplace. As IT budgets become tighter, the ability and desire of content developers to create content for both companies’ technologies is waning and RN understands that MMP may well become increasingly dominant because many perceive it as simply easier to use. Obviously, this is something that RN does not want to see happen and through its partnering with Intel, the company has created a strong channel by which to deliver its offering to the desktop independent of MS’s influence. Of course, history could have been different, if the two companies stayed in agreement with respect to file format, which was in unison in the version 4 days of their software.
Overall, this is an interesting development in an evolving space. Is RN trying to make sure that its brand is still visible to its current install base as it delivers an increasingly broader portfolio of services, or is something else afoot? Certainly, RN’s vision is that television is where the company wants to be long term and on the PC in the short run, perhaps as a bridging strategy or to hedge its television bet. But this vision, like so many others, rests on the fulfillment of TV and computing convergence — a convergence that is coming slower than some originally hoped. For the present, RN has developed an interesting strategy to solve its PC real estate problem. As a user, one would certainly want to understand how chip level upgrades of the player are supposed to come about, but this technology is no longer living in the rapid paced development frenzy of a few years ago. These technologies are relatively mature, so the potential impact of being several revisions out of sync with current product is lessened. We trust that folks at Intel and RN have at least considered this niggling issue — although painful memories of software overlays of firmware and hardware do come to mind. In the end, we believe RN will need to more clearly articulate its vision of the future, as the company’s fixation on television does not seem consistent with the actions implied by the Intel announcement. What would be better is if RN simply defined itself as a services-content provider regardless of access device and began delivering on that vision to its multitudinous clients. It would certainly help the market understand what is Real and what is not.
Hewlett Packard and Hughes Network Systems (HNS) announced this week a strategic alliance to jointly market a managed suite of content delivery solutions for enterprises. The new HNS offering will be available under the name DIRECWAY Content Delivery Services (CDS). The suite of services includes multicast streaming, multicast delivery, video on demand, and network acceleration that will support business applications such as real-time corporate communications, distance learning, and Internet back up. DIRECWAY CDS delivers bandwidth-intensive content via HNS’s satellite broadband network directly to an HP media cache appliance at each enterprise site. Corporate users can then access particular, preloaded content as needed. DIRECWAY CDS is scalable and comes packaged with IP multi-cast features including encryption, intelligent caching, and streaming compatibility with Real Networks, Windows Media, and Quicktime. Under the terms of the agreement, HNS joins the HP Service Provider Program (SPP) as a Select level member. The HP SPP is designed to accelerate the adoption of service offerings, and included features such as co-selling and co-marketing, joint solutions development, and end-to-end infrastructure, applications, and operations certification programs. No pricing or availability information was included in the announcement.
A study of broadband data access trends consistently shows satellite-based solutions consistently trailing the pack of mundane broadband technologies such as DSL and cable. The reason for this is simple. Until relatively recently, satellite solution providers such as Hughes primarily focused their attention and efforts on entertainment offerings, apparently seduced by Hollywood’s siren-like insistence that the big bucks always reside in consumers’ pocketbooks. But some funny things happened on the way to those supposedly easy pickings. Despite the industry’s best efforts, satellite access remained more complex and expensive than commonly available cable offerings, and though downloading Internet data from satellite can be blazingly fast, uploading data historically was far slower and problematic.
A funny thing happened in the business sector. As enterprises built increasingly robust WANs to support network traffic in widely dispersed campus environments, they often ended up with data environments that could easily support bandwidth intensive processes such as real-time video conferencing and distance learning. Where that data would originate from was the problem, which is where we see HNS and HP stepping into the picture. For HNS, DIRECWAY CDS simply leverages the company’s existing infrastructure to stream business-related data to specified sites, in much the same way as the company’s movies-on-demand feature works for consumers. We are particularly intrigued by the notion that CDS can support Internet backup, and wonder if the technology might be used to mirror data to remote sites for business continuity efforts. While we cannot speak to the demand for such services, we believe CDS could offer a reasonable return with very little risk for HNS. The offering should also bolster HP’s vision of “service centric computing” by enabling the delivery of enterprise data to virtually any location.
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