HP's Next CEO Needs To Steer The Ship

With Mark Hurd out of HP, the company has an opportunity to find a replacement who can carry the company forward. One of the striking moments at CiscoLive was listening to CEO John Chambers during the keynote and in a smaller gathering, and realizing how firmly he has his finger on the pulse of Cisco's business and technology direction. He is clearly driving the company. Whoever takes over HP, and apparently our own Art Wittmann interviewed  for the job, has a number of challenges facing them, not the least of which, as Bob Evans points out, is defining and articulating the HP data center and networking strategy.

One of those challenges is to integrate their IT divisions--servers, storage, networking, networking management and professional services--into a cohesive whole focused on selling HP branded equipment as a drop-in replacement for the mish-mash of equipment IT is running in data centers today. Cisco is finding out how difficult it is to break into the server market with UCS, but they are making headway. HP has yet to find out how hard it will be to break into the data center networking market with their own equipment, but every day they wait is another day lost.

Since HP acquired 3Com, completed in April 2010, we have been waiting to see what the company will do with its newly acquired networking gear. There's no question that HP had to acquire a switch vendor to have a glimmer of hope of competing in data center networking. The Procurve line, now called the "E Series," wasn't going to cut it. HP hasn't articulated to IT what the plans are for data center networking (other than throwing barbs at Cisco). Granted, the 3Com deal closed just a few months ago, but we are talking about a company who finally realized in 2009 that they had a switch line and started making blade switches for their chassis systems.

The question is, can HP really compete in the data center networking market? Brocade, Cisco and Juniper have all come forward with product road maps targeted at data center networking. All of those vendors are promoting new data center network designs that flatten the network, make more efficient use of the interconnects. What has HP done? Not much, other than say they are waiting or the data center bridging (DCB) and multi-pathing standards to be complete before supporting them.
 
Cisco is going full-steam ahead with product announcements like Overlay Transport Network for Layer 2 WAN connectivity, FabricPath for multi-path Ethernet in advance of standards, and support for Fibre Channel over Ethernet, which embody many of the feature requirements for the next generation data center network.

Juniper's data center play is more road map items, but the company's plans for treating multiple switches as a single logical unit sharing a single MAC table, their push for a single OS regardless of networking device and hitless upgrades are certainly pushing the envelope of data center switch features, even if all of them aren't shipping today.Meanwhile, HP beats the standards drum and falls further behind. Standards are wonderful things, but HP needs to shake up the market and get some attention by talking about the "A Series," composed of the H3C switches acquired from 3Com, and what it can do for data center networking. Cisco is market leader and the company to beat. Cisco is also the company being talked about most. HP has eeked out some discussion about FlexFabric, which attempts to simplify cabling, but these are incremental issues that won't get IT interested in HP as a networking company. 

The next few years in data center networking are going to be about flatter, faster, robust networking, carrying data and storage traffic and supporting virtualization strategies. HP owns most of the components for a full data center with the exception of FC switching, which it OEMs from Cisco and Brocade. The company just needs to plan, articulate and execute on a holistic data center strategy.

Finally, their professional services arm, HP Enterprise Services, which Paul McDougall says hasn't been used well, needs to be clearly focused on moving existing customers off Cisco infrastructure and onto HP gear. That is going to be a difficult process and some customers won't migrate, but if HP can offer a better network architecture with clear capital and operational savings, HP can start winning deals and increasing market share.