I'm not sure whether the Compaq merger is far enough in the past for Hewlett-Packard to contemplate another major acquisition. But if so, I believe that HP could really shake up with networking status quo by buying Juniper.
This move would certainly take some chutzpah, and it would cost HP somewhere in the neighborhood of $30 billion to pull this off. Nevertheless, I believe that an acquisition of Juniper makes a ton of sense.
First off, HP and Juniper have almost no product overlap. HP has a strong business in Ethernet edge and core switches, as well as lots of network management strength with OpenView and now Opsware. Juniper, by contrast, is a leader in routing, security and application networking. A combined entity would offer top-notch edge-to-core network infrastructure, along with all the goodies needed to secure and manage the whole enchilada. In other words, HP would be able to pretty much match Cisco Systems product for product.
Would enterprise users actually buy networking equipment from HP/Juniper? Yup. Cisco owns the enterprise network market and there isn't a No. 2 vendor in sight. Viable vendors that establish this position tend to do extremely well. Remember Amdahl and HDS in the mainframe market? These guys gave IBM fits in the data center for years. And it isn't like HP is a stranger to the enterprise.
Finally, this new networking tide could float a lot of other HP boats, pulling in server, PC, software, and professional services dollars as well. HP would also bolster its already strong presence in the global telecommunications sector, an industry that is growing quite well again.
Lots of system vendors like IBM and Sun Microsystems walked away from networking, and Cisco has benefited every time. Quietly, HP has built a multibillion-dollar, successful business with its ProCurve networking division.
A Juniper acquisition would be an extremely bold step. Although HP may not want to stick its neck out quite this far, I believe there is a potential huge payoff if it does.