It's not that Cisco's move into servers isn't an important one; it undoubtedly is. Its notion that servers should be less about a pile of sheet metal housing processors, memory, storage and expansion slots, and more about tightly packed processors and memory with a big fat pipe that readily connects to similar servers, is right on the money. Fully embracing virtualization and building servers that work best in a virtualized environment is a game where Cisco can compete. But it's just not as important as driving a software infrastructure that can support applications running securely and reliably anywhere and on any platform.
The transition to the data center that Cisco imagines will happen, but it's not a rip-and-replace imperative. As servers and networking gear reach amortization, smart data center architects will buy new stuff that will look more like what Cisco described last week, but whether they buy it from Cisco is another matter. IBM and other server vendors are used to dealing in the razor-thin margins that Cisco has never tolerated, and not even Cisco would create an environment that works well only with its own switches and servers.
This move also is a good sign for economy watchers. IBM reportedly is discussing paying a healthy premium for Sun, implying that it thinks Sun is worth far more than its market cap would suggest. As other tech companies with cash on hand start to feel comfortable that they know the scope of the recession, it's likely we'll see a lot more consolidation, a sign we've found a market bottom. If there's bad news for anyone in last week's moves, it's for Brocade and anyone else betting on the long-term viability of native Fibre Channel. Unless someone like EMC ponies up for Brocade, you can bet that native Fibre Channel will be just a memory in five years.
Register to see all reports at InformationWeekAnalytics.com.