What is the Right Investment Angle on Data Center Construction?

Okay, Microsoft’s Ray Ozzie comments at the Goldman Sachs technology investment conference in Las Vegas today got me thinking: How do you invest around data center construction? Because here is Ozzie:

In order to continue to seek advantage over Google in online services, Ozzie said that Microsoft will continue to invest in massive data centers, such as its $550 million center recently slated for construction in San Antonio, Texas.

“There should be a recognition that demand exceeds supply with data centers, that’s why we need to keep building them,” Ozzie said.

So, what is it? Commodity servers? Ick. Power? Tough. Rack providers? Storage providers? Hmmm. I’m open to ideas.


  1. 1. Huawei for networking
    2. VMware (upcoming IPO)
    3. Thermal Management sytems (no companies yet) for using exergy (not energy) within data centers more efficiently


  3. Data center infrastructure is a race to the bottom – servers, switching, etc. Price pressure and shrinking margins make this a rough patch to invest in.
    I think that the best place to invest would be in systems to help the customer milk the most out of the infrastructure that they have invested in – maximizing operational efficiencies – that would seem to be a good place to be in my mind.
    Opsware type investements……

  4. Boy – this theme has bubble written all over it.
    Remember the former Intel data centers, all the leftover Exodus/C&W garbage, a bunch of more from Sprint? All these guys pumped billions in the last build out frenzy (with gazillion $ storage arrays and Sun servers) and drove their trains right into the wall of 2002/3. They were unloaded for essentially zero by 2004. u-g-l-y.
    And did someone say Opsware? Remember these guys were LoudCloud – purely built on the managing data centers. Even the enterprise software is walking dead now too.
    The closest thing to opportunity there are raised-floor and creative HVAC vendors.

  5. Nah, there’s some opportunity, although I may be slightly biased. 😉
    The operational guys like Opsware have some room I think. (Yes, the former LoudCloud, but hey, if startups don’t change business plans/models, are they really startups? And I’m biased because I had some friends and associates there.) Squeezing efficiencies out of all that commodiity hardware becomes more critical every day. I see it with my current job, working for a large datacenter company (again). What I wouldn’t give for Opsware in my day job.
    I’m a believer in storage, esp. online and backup. Of course, I used to work for a backup company that EMC recently purchased. (Like I said, biased.) But storage isn’t decreasing; it almost can’t. For all the things that will change, the need for more storage isn’t going anywhere. Neither is the need for more data protection in various forms. Now, what form that investment takes becomes a bottom-up analysis, not a top-down one IMO.
    My $0.0000002 (adjusted for inflation)

  6. anonymouse says:

    The data center segment for any company directly investing is a pure play drain-ware. Extremely high capital outlays coupled with depreciation and little to no return on investment (a certain L3 comes to mind).
    I would outsource to contractors and just pay a monthly outgoing on a multiple hub and spoke model littered all over the country and keep just the application management in control.

  7. invest in real estate in regions with abundant power supplies, particularly hydro-electric or other green sources

  8. bashiru sule says:

    i want to really be rich even though am in Africa how can do make that possible? there are alot of resourses but no planning hence the resourses go wasteful. My aim strategically improve as if am in America europe or asia