Forced to pair up by the growing hardware and software dominance of IBM and Oracle, Microsoft and HP could grow increasingly closer over the next couple of years.
Among the flood of computing vendor announcements that marked the start of the new year was one that might have been easily overlooked: Microsoft and HP unveiled converged application appliances for small businesses, business intelligence, analytics and email serving.
Neither Microsoft (NASDAQ: MSFT) nor HP (NYSE: HPQ) is the first to provide "turnkey black boxes" for business intelligence, and neither Microsoft nor HP has been as front-and-center in press and analyst coverage as Apple, Google, or Facebook over the last two to three years — nor, apparently, as financially successful. Moreover, on the face of it, nothing suggests that this is anything more than the latest in a long string of vendor announcements that do not indicate partiality for the partner du jour on either side, but rather the fleshing out of each side's alliances of the moment.
I would suggest, however, that this may — may — be an indicator of a much bigger computing industry realignment, with major implications for IT and enterprise application buyers.
Setting the Stage: Industry Platforms Simplify
It appears that, over the last few years, users' choices between vendor architectures have become significantly less complex. The folding of Sun into Oracle (NASDAQ: ORCL), as well as smaller acquisitions of second-tier database vendors (Sybase) and middleware and enterprise application suppliers (from PeopleSoft to BEA), means that a few major vendors now clearly outstrip the rest in covering the entire software and hardware stack. It is, in fact, possible to say that there are three major choices for "black boxes for all needs," of which two are Oracle and IBM (NYSE: IBM).
At the same time, the choice of platform becomes ever clearer. Two "movements" have survived and thrived in the chaos of Internet innovation: the Windows platform and the Linux one. Moreover, two scalability technologies have become dominant — "distributed"/scale-out and scale-up — both of which have added strong virtualization capabilities over the last decade, and are therefore well positioned for cloud computing and "cloudified," virtualized enterprise applications.
Fairly or unfairly, IBM is seen as stronger in scale-up and less strong in scale-out. The question then becomes, who will dominate the business market for scale-out platforms? What vendor or vendors have the experience, breadth of products and services and ability to bridge Windows and Linux that such a dominant position requires?
Oracle is an obvious candidate. It now has the ability to integrate scale-out hardware, storage, middleware and infrastructure software (especially for data management) that lets it fit scale-out environments like PC-dominated large-enterprise divisions and medium-sized businesses, as well as resuscitate the remains of the "Oracle and a Sun" Internet startups of the late 1990s, and use the same platform to fit all kinds of popular architectures, ranging from grid to loosely-coupled to public-cloud server farm. And, of course, it has a strong enterprise app suite and BI offering.
However, there are two problems. First, Oracle has until recently consistently been sending the message of not supporting the kind of development and database anarchy that has characterized MySQL, Hadoop and Vertica. And second, Sun hardware revenues and market share, long declining, appear to be continuing to shrink — meaning that Sun/Oracle hardware is becoming less of a factor in public and private clouds and enterprise app platforms.
Wayne Kernochan of Infostructure Associates has been an IT industry analyst focused on infrastructure software for more than 20 years.
Microsoft and HP: Better Together Than Alone
It's easy to look at the rapid revenue and market share gains of Apple, Google and Facebook and dismiss both Microsoft and HP. However, from a platform perspective, both Microsoft and HP have been surprisingly successful. HP's acquisitions — especially that of Compaq, which has shifted HP's center of gravity to Intel chips and PC form factors — have allowed HP to add strength in scale-out to strength in scale-up, and to "catch the wave" of bottom-up PC penetration of data centers. The result is that over the last two decades, HP has gone from a small fraction of the size of IBM to about 30 percent bigger in revenues — and the gap is widening. Note that the majority of those HP revenues come from PC form factor hardware and PC-market printers.
Meanwhile, the last 10 years have not been pleasant ones for Microsoft, and especially because user perceptions have been that its software is increasingly less relevant in the smartphone era. However, the company has been on something of a roll as of late, with the amazing consumer success of Kinect in the last few months and two straight quarters of greater than 20 percent revenue growth, and several commentators note that Microsoft appears to have finally won the business office suite wars except in some parts of Europe. In addition, while all eyes have been on Microsoft's efforts in the consumer and Internet arenas, it has managed to leverage its scale-out PC-server-farm approach into a ubiquitous position in many large enterprise data centers and a dominant position in many medium-sized firms. In these, Office, Outlook, SharePoint and SQL Server actually run the business, with Microsoft-based BI a significant presence in its customers' shops; and Microsoft has the business services to support them. Finally, while Microsoft is not yet a Linux enthusiast, its relationship with that movement has clearly lost much of its edginess.
The result is that both Microsoft and HP have strong offerings for inclusion in scale-out platforms. No one dominates Linux, but HP has an exceptionally long history of supporting Unix and then Linux at both the small server and data center levels. Meanwhile, Microsoft is dominant in the Wintel business market, and has used its clout to move Windows into ever-higher levels of its customers' IT and analytical processing. And both HP and Microsoft have a hold on business spend driven by the consumer market that neither IBM nor Oracle can seriously dent.
On the other hand, neither HP nor Microsoft alone has the breadth in the scale-out space that IBM does in scale-up. HP lacks infrastructure software; Microsoft lacks Linux and hardware experience. As the market moves to specialized appliances for BI, ERP, and so on, in which distributed hardware and software can be integrated and fine-tuned from the bottom to the top of the storage/hardware/software stack for such uses as BI, these shortcomings become more important.
However, combine Microsoft and HP — call it MicroHP — and what do you have? A full Windows-plus-Linux scale-out hardware and software lineup, with an exceptionally strong position both in SaaS/public cloud and data centers, and a huge presence on the business desktop. This would allow such a combined entity to produce well-tuned appliances for such hot areas as BI/analytics — just as, apparently, Microsoft and HP have just done.
Such a combination does not have to be formal. Indeed, considering the massive egos prevalent in the computer industry, it is hard to imagine either side considering a merger. However, that does not prevent IT buyers from treating the two as one.
What Does MicroHP Mean for IT Buyers?
At this point, some readers will say that a concept like MicroHP is the result of the analyst consuming too many microbrews. However, those IT buyers still willing to stay with me may want to consider the following thoughts and action items:
- Over the next two years, the evolution of cloud computing is likely to drive Microsoft and HP closer together rather than further apart. Microsoft can't count on the consumer market to keep its revenues growing; it will need to improve its business offerings, not just use its consumer design teams to drive new business products. Meanwhile, HP faces stronger competition from both Oracle and IBM, competition in which Oracle and IBM will point to the greater degree of integration of their software and hardware that can lead to greater performance — so HP will need a better software integration story, preferably one that emphasizes Windows as well as Linux.
- IT BI buyers should consider making one entry on their short lists read not HP or Microsoft, but HP/Microsoft. By asking both vendors to act as one, you will send a strong message to them that should result in greater efforts at integrating their product lines, and therefore better products for IT private/hybrid cloud or analytics needs.
- IT buyers should also ask for clearer roadmaps on "Phase 2" of cloud deployment. Many of the initial private cloud deployments have been for Linux, while Windows deployments are often independently done. HP and Microsoft should be able to find a way to help customers integrate the two cloud platforms better.
One final note: when I first entered the computer industry in the 1970s, it was hard to imagine a day in which another vendor would have greater revenues than IBM. I have lived to see that day; and now I am imagining a firm that would be twice IBM's size and growing faster — and yet it would be relatively invisible compared to hot firms like Google and Facebook that were created a decade or less ago. The industry remains, after all these years, never dull.