The news broke about an hour ago that HP is buying 3Com, leading to a glut of theories that Cisco is under some tremendous new threat. If that is the case then Cisco really should have picked up 3Com or some other datacentre networking concern many moons ago.
HP buying 3Com might sound like a combination from the days when the Reagan/Thatcher axis (and even more dangerous Wham axis) was still intact but it does makes some sense, providing the former with a slingshot to attack core networking deals that Cisco tends to gobble up without breaking sweat or margins. But Cisco remains the Goliath here and Davids rarely win.*
I think that what HP is doing here is rather less interesting than mounting a credible challenge to Cisco; it is adding accretive revenue and taking advantages of economy of scale just as it did with EDS and a ton of other previous deals. If, as some pundits suggest, Oracle is the new CA when it comes to M&A, then so is HP (and IBM and Symantec, and others).
HP is picking up 3Com at a price that today constitutes a mid-priced deal. Yes, it complements HP's rising capabilites at the network edge and provides network security tools and inroads to China but Cisco is part of the furniture at the core. The main significance of this deal may be to spur other, more significant transactions, especially those that unite networking with datacentre infrastructure, the most obvious being Cisco-EMC. Now that really would be a trumping.
One final thought: if HP had made this deal a little over a decade ago it could have picked up Palm too as part of the package -- a sort of twofer purchase that could have changed the whole shape of the market.
*Although sometimes of course they do, as I found to my cost betting on David Haye against Valuev in the heavyweight boxing bout at the weekend -- adapt view of above based on this data if you like.