September 15, 2010
The other day I was just reminiscing with a friend who works at HP about all the good times I had there with my ProCurve family. When I left for a once in lifetime opportunity, I had so much hope for HP’s networking division. Like many of my inquiries from global customers looking for a Cisco alternative, I’m concerned about the division and its long-term viability. I’m not worried if HP will continue to exist without Mark Hurd. Companies are more than a single leader. There is plenty of research, books, and online debates about the effect of a single person: Jack Welch, Steve Jobs, John Chambers, etc. The issue at hand is the existence of product lines within enormous companies, like networking within HP. One of my mentors always said, “If you look at networking over the last twenty years, no major IT company or voice vendor has been able to pull off being a serious networking vendor if networking wasn’t its first priority.” Fundamentally networking is one of the few technologies where a vendor has to be all in. The networking graveyard is full of headstones: Nortel fell off the face of earth, IBM sold off its assets, and Dell hobbles along.
Ah, you might say, what about HP? That brings me to my three observations that every IT manager should consider when including HP in their network architecture:
- Observation #1: Truth be told, HP just got into networking. Over the last 15+ years, the division has waxed and waned with each CEO depending on their strategy for HP. Lewis Platt, Carly Fiorina, and Mark Hurd were ambivalent, scorned, or loved networking. For the vast majority of those years, HP and HP Services sold Cisco instead of its own networking products. Luckily under John McHugh’s tenure during HP’s CEO revolving door, he grew the division from $100M to $1B by convincing Palo Alto that the division should be run as a separate organization with its own brand, ProCurve Networking. Until a few years into Hurd’s term, HP’s networking division acted as a separate networking company and flourished.
- Observation #2: It took a visionary CEO to catalyze HP networking (pun intended). Hurd was the first CEO to see value in IT plumbing, and it was more than the revenue and margins Cisco dominates. He believed HP’s data center push and dominance were dead in the water until it had viable data center switches that provided it with strategy and solution differentiation from Cisco and IBM. Early on, he carved out extra budget for ProCurve to start developing data center switches, and when R&D wasn’t quick enough, he went after 3Com and brought Dave Donetelli who spent 22 years at EMC and emerged as the premier thought leader in data center convergence and virtualization. In the midst of bringing networking back into the HP fold, the networking division has floundered around the $1B mark for the past two years.
- Observation #3: HP’s next leader needs to push the networking group beyond being commodity gear. The question is will the next CEO have the same vision, drive, and clarity to invest in networking? HP needs areas that can grow revenue and margins to offset the erosion of PCs, printers, and black gold. With owning only a portion of the servers, storage, and services market, can HP afford to move research, sales, and marketing dollars over to networking? The division only brings in 1-2% of HP’s revenue — even with the 3Com acquisition. Networking is at another pivotal point that requires a lot of R&D money. HP needs to increase its networking service and support along with its sales force because networks are still being analyzed, specified, deployed, managed, and upgraded by networking professionals. HP will need to quickly come out with a vision, strategy, and road map that has more meat than its current message: commodity switching and data center consolidation.
Networks are a 5-8 year investment, and there are too many changes and unknowns to make large investment into HP’s networking division. Infrastructure and operational personnel should not run away from HP but be cautious and take edge to core approach to refreshing the network. Trends show companies moving to a dual networking environment, and the market is offering solid networking alternatives like Juniper and Brocade.
In the meantime, while we wait for HP, we have the OpenFlow movement, mobility, empowerment, virtualization, and data center convergence providing ripe opportunities to shake up the networking industry while we watch HP and its Minnow weather the latest storm. Speaking of data center convergence, keep your eyes peeled for reports from me and Rich Fichera about data center, virtualization, data center networking, and data center convergence. Leverage James Staten’s principles in his report about virtualization and data center consolidation, Are Converged Infrastructures Good For IT, to prepare your next network refresh.