Opportunistic Computing in the Cloud
by
Charlie Bess
I've been thinking more about the entry the other day on heeding Moore's law.
I've had a few conversations with folks in other organizations about cloud computing. I am always fascinated by their almost blind focus on cost reduction. I see real opportunity for the shift in the value side. Value is what makes IT relevant. An exclusively cost focus is why IT doesn't matter.
Some visionary development organizations are starting to see a future where cloud computing is the default and stand alone applications the exception. Even the desktop will be a little cloud of computing before the end of the decade. Once the development organizations/industry start new projects from a cloud perspective, the concept of the data center will be forced to shift. After all, the only reason for all the hardware is to derive business value from the apps running on it. Once the apps move on, the infrastructure will need to as well. Virtualization is great, but it is only a step along this road.
With our constrained computing perspective, we significantly limit how we think about delivering value. We need to shift our thinking when we move into a computing "age of abundance". When software development turns into assembly instead of creation, maintenance of the code may have more to do with simulation than with what we think of today as maintenance. I wonder when ITIL will incorporate this concept? The ability to understand information flow between components and rapid adjustment to business needs without having business impact will define the quality maintenance organization of the future. Unfortunately, I've only started to see universities preparing individuals for this vision.
Once we have the opportunity to access an almost unlimited computing resource, what we will use it for will shift. We can move to a cause and effect perspective, where we recognize patterns and likely outcomes and adjust to have the future we want rather than respond to the one we get. These different perspectives of the problem are the difference between the way humans who play chess well and the way most computer programs play chess. In this cloud computing age, we should be able to have the best of both worlds. Unlimited computing to recognize patterns in the massive amount of fine grained information the advances in sensors and edge computing will provide, along with the parallel processing that can be applied to looking at the implications of the likely outcomes.
Microsoft has also done some interesting work with knowledge network to tie people together that work on similar problems, across a large organization. This type of project is a starting point for what you'd do with an unconstrained view of computing. Using the capabilities based on the opportunity to derive knowledge from the information flow, not just in the process itself.
