This might surprise you folks, but Smart Grid (intelligent power distribution and management) is even more important to our nation’s future success than the Cloud is! Interestingly enough in many cases one enables the other or at least dramatically improves its usability and performance.

The current situation in our data centers (Infrastructure 1.0) has many things in common with our current electrical grid (lets call it Grid 1.0). The IT infrastructure in our data centers today is very manual, use isn’t easily measured and demand isn’t easily provisioned or de-provisioned as requirements change. The infrastructure is managed through a combination of tools and spreadsheets that require on-going support from IT staff. As pointed out by other members of the Infrastructure 2.0 group, we’ve created highly qualified “IT Clerk” positions in order to support this very manual and complex issue. By its very design we’ve created an environment that’s much like a hologram, it works great when you’re in your nice little box (four walls of the data center), but don’t try to get the hologram to step outside or poof, no more hologram.

So how does Infrastructure 1.0 relate to today’s electrical grid? Like the inflexible data centers of today use of power and the demand management of power are fairly complex and often times manual processes. We have a limited ability to dynamically allocate power to “important” functions like hospitals or the fridge in your home. We don’t have visibility up and down the chain of delivery to determine where and how the power is being used and whether its delivery is as efficient as possible. Like your current data center (assuming it’s not fully virtualized on VMware vSphere) you don’t have a silver bullet solution that can help you manage the available resources efficiently. A good example would be the ability to use idle or dev systems when demand increases. Today when your application reaches peak demand you need to have all the hardware sitting waiting for the demand or you’re going to have angry customers as the system slows or fails. In the perfect world your systems combine as one larger resource and when demand increases for that “resource” your platform can automatically allocate additional capacity from idle systems. Grid 1.0 is very much the same in that it’s not easy to manage demand and power companies have to over provision or buy very expensive “resource” from other areas.

At this point you still might be asking yourself, “how does this have anything to do with the cloud”, well, I’ll explain.

When a cloud owner needs more capacity they add resource to the environment, regardless of whether that capacity requirement is storage, memory, network, or CPU. In a Grid 2.0 (Smart Grid) environment the same can be true as power resources are generated the new capacity could be dynamically allocated to where it provides the biggest benefit in price and performance. In Grid 1.0 often times power generation capacity goes unused even when a region is under stress because the overall grid can’t easily redistribute that power outside of a fixed area. One example would be putting thousands of windmills in the middle of the country, sounds great right, cheap land, few people to disturb and lots of wind. Unfortunately the only people the generated power could serve would be those within a limited area around the windmills.

So why is the cloud impacted by the generation of power and how can it help the power grid?

If you’ve got a “Smart” cloud you could dynamically manage your capacity as it relates to power, just the same way that you could when user demand changes. If your local utility needed more power to manage peak load your environment could adjust to reduce power use automatically based on an electronic request from the provider. Your cloud could also move capacity all or in part to an alternate location where power availability isn’t an issue. This capability provides both the customer and the utility with many benefits not the least of which is saving on the bottom line. The issues pointed out here all point to the greater need to understand how we’re using our valuable resources, whether that resource is a CPU or energy.

I believe that the future strength of our economy and even the ability to maintain any level of growth will depend on successful adoption of Grid 2.0, the Cloud and Infrastructure 2.0. If we don’t solve the problem of Grid 1.0, much of the benefit associated with the cloud won’t be realized. If we don’t develop Infrastructure 2.0, we won’t be able to fully benefit from Grid 2.0.  We have a growing population, which translates to a growing need for power and compute resource. Unfortunately, this situation of ever increasing demand, coupled with the limited ability to dynamically distribute or allocate energy resources means we eventually hit a wall. In 10, or maybe 20 years we will reach a point where we’ll have to make painful decisions about where power plants are built and we all know the answer “not in my back yard”.