Amazon’s Cloud Play vs. Google/Microsoft’s

I heard the figure of $100M in AWS (Amazon Web Services) revenues bandied about lately.  Not too shabby for a business that started out as a spare-time as-is reuse of Amazon’s infrastructure.  By far the healthiest supplier of IaaS (Infrastructure as a Service)

Compare them to Google and Microsoft, who started PaaS (Platform-as-a-service) businesses instead of IaaS.  PaaS (if I understand it right) tries to get developers to develop applications for the platform, whereas IaaS “just” provides a run-time environment for applications built elsewhere.

Microsoft of advertising the bajeezus out of Azure, but I’m not hearing much about great results.  Certainly no $100M in revenues (yet?).  And Google seems to be flailing about with its various cloud and app-dev platforms.

Makes IaaS look like the right bet, at least so far.



2 comments on “Amazon’s Cloud Play vs. Google/Microsoft’s

  1. Anonymous says:

    In my opinion, there are two aspects to the IaaS vs. PaaS vs. SaaS discussion: a) Strategy and b) Technology.

    a) Strategy: From a strategic point of view, “the cloud” is an important building block for Microsoft to achieve its goal of retaining its leading position in desktop and server technologies. MS Office, Windows, Exchange, SQL Server and SharePoint are Microsoft’s big cash cows that need to be protected at all costs.

    To keep its cash cows healthy, Microsoft is betting the farm on two different cloud offerings:

    Office 365 (SaaS): Microsoft is offering its familiar MS Office suite, combined with its three essential collaboration servers (Exchange, SharePoint and Lync), within the Office 365 package. The Office 365 offering caters to companies that are not willing to maintain several hardware-hungry in-house server products and that would otherwise go to Google Applications.

    Azure (PaaS and a little bit of IaaS): Azure is Microsoft’s ultimate effort to get the enterprise entirely hooked on Microsoft’s core technologies: .NET, SQL Server, SharePoint Services, and so on. If you buy into Azure, you get some great advantages (listed below), but you also limit your choices when it comes to integration with non-Microsoft solutions. To add a little more flexibility, Microsoft is now allowing the use of raw Windows 2008 R2 VMs on the Azure platform. In other words, as long as you stay with Microsoft technologies, you get your IaaS offering.

    b) Technology: From a technical point of view, IaaS is a lot easier to implement, but not as powerful as PaaS. Therefore, IaaS might be an interim solution to rid your in-house data center of the most taxing servers, without having to rewrite your applications. The migration is simple, you have full server-level access and there are minimal training requirements for your IT staff. On the other hand, you cannot significantly reduce your IT staff, as the IaaS servers require almost as much maintenance and upgrades as if they were located at your own data center.

    Once your applications have reached their end of life, the advantages of Microsoft’s PaaS model may sway you towards writing the new software for Azure. Here they are:
    a. Automated system patches
    b. No server upgrades required
    c. Scalability
    d. Marketplace where you can buy reusable modules

    Long story short, IaaS providers like Amazon, Rackspace or GoGrid make it easy to quickly eliminate datacenter availability, maintenance, scalability and performance issues. I think that this is why we are seeing pretty nice revenue figures from those providers. But when we look at it from a profitability perspective, I would imagine that the Microsoft PaaS model (with a little IaaS thrown in) is superior. Whereas the IaaS model commoditizes hardware resources, PaaS offers pure services, making it more scalable while at the same time pushing Microsoft’s cash cow technologies. And for the actual customer the PaaS model seems to have a great value proposition, as it eliminates some of today’s central data center headaches (patches, upgrades, server provisioning, and so on).

  2. Dan Gordon says:

    I’m told by more than one source that my $100M figure was for one quarter, not a year. the yearly figure may be in excess in $500M for 2010

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s