Sometimes we are so focused on what the (technical) media tells us that we can lose perspective on reality and market dynamics that can (and often do) affect our businesses. Analysts are great and perform a valuable role, but they are not entrepreneurs, inventors or architects. This being the case, if we put too much stock in what they say or where the market will go, we can mislead ourselves. This could result in a false sense of direction leading us to the wrong conclusions about our roadmaps. Such is the case with Cloud Computing and mass delivery of Internet Protocol applications and services.
To set the stage let me start with the misconception that Cloud Computing has somehow been isolated to large ISPs such as Google and Amazon. We could include Microsoft in the mix but they are currently not offering a consumable solution yet. Why would we believe that these companies are the only focus of the market? What about the Fortune 100 communications companies that have already been in business for many years providing IP communications and services in mass? Isn’t it possible that these are the “real” contenders in the Cloud Computing space and we keep listening to analysts tell us ‘they don’t belong’ or large service providers are in a different segment and therefore aren’t on their radar.
I have to admit I’ve found this to be the case. Within the large analyst firms there are specializations of expertise; IT, Virtualization, Infrastructure, Networks, and Communications to name a few. These analysts often don’t communicate or collaborate with each other and as a result we get a very monolithic view of an industry. Since IT has evolved to be in every sector from Energy to Entertainment, gaining a holistic view of future offerings (Cloud Computing included) is difficult to say the least when confronting “experts” in individual sectors. All the while huge communications and entertainment companies such as AT&T, Verizon, Deutsche Telekom, and Comcast are building out massive IP networks and services that would dwarf any Enterprise, ISP or Software company.
Note that the convergence of IT services for a “utility” offering is the core definition of Cloud Computing and yet Utility type companies are not usually included as contenders in the space. I would venture to guess that companies such as AT&T for instance should be considered players in the Cloud Computing business. Wouldn’t you? To prove the point, one only has to look at the offerings available in the market today and see that there is a huge potential to fulfill the promise (and demand) of Cloud Computing.
The diagram below is a depiction of the current AT&T U-verse network. Note that it is a pure Internet Protocol (IP) network and provides utility based data and communications to a mass population with Service Level Agreements (SLAs) capable of providing High Definition Video Entertainment services, IP Voice services, and Internet Applications services. AT&T has two major data centers which aggregate traffic from Video Head Ends as well as their own ISP connections and contracts to Internet asset providers such as Yahoo!. The IP traffic flows from the data center via fiber optic cable to Digital Subscriber Line Access Multiplexers (DSLAM) within a neighborhood or office complex. These DSLAMs then provide network access to the individual home or office via Very High Speed Digital Subscriber Lines (VHDSL) capable of carrying high bandwidth IP applications such as HDTV or heavily intensive graphics applications. VHDSL is an asynchronous network technology meaning there is a disparity between upstream and downstream bandwidth. However, the upstream path is still very capable of providing certain types of real-time communication and high bandwidth IP data/applications paths.
Further, if we take a look at the AT&T U-verse home/office extension we’ll note that Cisco has partnered with AT&T to provide a low cost, high speed (100 GB) combination wireline/wireless router. They have also created an extended end point Set Top Box (STB) which is really a Central Processing Unit (CPU) with both volatile and non volatile memory. Further, Microsoft is included in this partnership and has written a light version of their Operating System (O/S) called Mediaroom in order to provide a platform for services from the data center. These include custom applications that can be accessed from any PC/Mac in the world and used to control both Entertainment and Communications applications rendered either on the IPTV or on the local PC. Mediaroom resides on the STB and communicates to the data center through the Cisco router using IP. Microsoft has installations of Mediaroom in (25) major service providers around the world in every geography (including China).
Applications can be on-boarded through this network into the home/office by merely sending the code over the IP network to each individual STB in the home/office. One has to wonder will it be long before AT&T, partnering with Microsoft and Cisco begins to offer applications for home/office use over this network as well. All of the elements are already in place, being used and a billing model producing revenues. This is not the case with Google, Amazon or any other emerging Cloud provider.
Similar types of services are available from other large communications companies around the world but they do not seem to be captured as companies worthy of being on the Cloud Computing radar. Cloud Computing will evolve that is for sure and many companies will either opt out or be pushed to the side by the real contenders. However, the convergence of IP technologies and the emergence of utility based computing is much broader than just a few ISPs and the sooner all the players are brought into the mix, the better for those of us who are attempting to set future roadmaps to enable the Cloud to be real.