The Case for HP as the Leader in Converged Infrastructure

When defining taxonomy for measuring segments in the IT space, I believe that there should be a corollary from Heisenberg’s uncertainty principal:

Any forecast of a hot topic will undercount the market as everyone jumps on the trend.

Converged infrastructure is not a new topic, with blade servers arguably starting the trend a decade ago, but one that is definitely in a hot part of the hype cycle. When Wikibon put forth the industry’s first market forecast for converged infrastructure last year, we had a healthy internal debate as to what differentiates converged solutions from solutions such as IBM Redbooks and EMC Proven Solutions that existed before. While customers are pitched a broad spectrum of options, at the end of the day, customers don’t consume infrastructure piecemeal; a vendor, SI, VAR or IT staff puts together a solution that is then utilized. The boundary between what is and isn’t converged infrastructure is therefore a chalk line along the value chain. HP can stake claim to popularizing the term “converged infrastructure”, and is the only vendor that is top 3 in revenue for servers, storage and network, so could HP be the leader in convergence?

the-transformation-of-HP-infographic.png

Wikibon’s 2011 Infographic on the Transformation of HP

Starting with blade servers, HP long ago dethroned IBM and despite rapid growth by Cisco, HP remains the leader. Blade servers primarily focus on providing dense compute (server) with basic networking capabilities; storage is typically attached through the networking options. At HP’s analyst summit earlier this year, it was clear that every group in HP’s enterprise group takes the convergence messaging to heart and is pushing to blur the lines between its product lines wherever possible. In recent years, HP has also cross-pollinated storage with compute – allowing storage modules to go into blade server chassis. The architecture of HP’s Moonshot also allows for a wide variety of compute, storage and networking modules.

So, if convergence is the baseline of HP’s product line, why doesn’t HP end up at the top of revenue charts for converged solutions? Current definitions define convergence of full stack solutions that include server, storage, networking and management. HP’s full stack offerings are VirtualSystem, AppSystem and CloudSystem. Perhaps HP has marketed itself out of the definition of some analyst firms. If the offerings that combine compute and storage were marketed as part of the xxxxSystem families, the combined revenue would likely surpass that of VCE (which reported its first quarter over $250M in Q4 2012) and the other players.

While HP has many challenges including a shrinking PC market, transition from older product lines to newer (such as storage from EVA to 3PAR), and low software revenues (3% of the company), it is one of the stalwarts of enterprise infrastructure with a large install base and solid portfolio. As virtualization and cloud computing increase the adoption of solutions that eliminate the boundaries between silos, HP needs to push its converged offerings towards emerging markets including OpenStack, various virtualization offerings and service provider environments.

 

Data points: (IDC public data, $M and rank)

HP server revenue 1Q13 $2,947 (#1), 4Q12 $3,626 (#2)

HP storage revenue 4Q12 $1,380 (Tied #2)

 

In the video below, I discuss Hyperscale Invades the Enterprise and Impact on Converged Infrastructure

Share
  • Bobby

    HP can’t compete with Unified Fabric/CNA/Fcoe/Ucsm/dcnm/prime

  • http://blogstu.wordpress.com stu

    Hi “Bobby” (would be great if you are transparent as to who you work for) – thanks for bringing up the point that HP leverages merchant silicon and leading components rather than mostly in-house solutions. There is no doubt that Cisco is a strong competitor with a lot of focus on driving converged solutions. Cisco relies on partners for storage. Networking margins have historically been higher than either storage or servers.

  • returnofthemus

    Whilst respecting HP and the Moonshot stuff, very impressive, but noisy, I think when the market eventually matures, they’ll end up a close second to IBM, not because of inferior technology, but because when you’re talking about application optimised systems there is no one on the planet that does it better. In fact when I hear the term Converged Infrastructure, I think Mainframe, what’s the difference?

  • Pingback: HP, Dell Angle for the Top Spot In Converged Storage : It’s a Race | SiliconANGLE