Energy Modeling I – The iPhone Challenge?

In the 1980s lots of time, money, and effort went into end-use modeling for electric and gas utilities in California and elsewhere. These models seemed to have some value in the residential and commercial sectors for examining the impact of building and equipment efficiency standards, and the consequences of changing penetration of appliances in households and businesses. They require a crazy amount of input data which is almost impossible to accurately obtain, and produce reams of output data which is equally opaque. End-use models are still around today, but don’t have much impact, probably because they are so difficult to use and understand. About the only real innovation in 30 years, if you want to call it that,  is adding in CO2 output for those who are interested in greenhouse gases. Hard to think of a less likely candidate for an iPhone application.

So let’s consider it. The programming tools are radically better, the user interface paradigm is radically different, and the iPhone itself is somewhat of an improvement on 1980s technology.

Typical time to run an end-use model forecast of 20 years in the 1980s was on the order of 1 hour of CPU time on an IBM System/370 model 3081 mainframe. That 3081 had a pair of processors with a cycle time of 26 nanoseconds — about 38.5 mHz — with up to 32MB of main memory. The iPhone 3G S has an ARM processor running at 600 mHz, along with a powerful GPU, and 256MB of RAM. So processing capacity probably won’t be an issue.

Power consumption? The 3081, depending on the model, drew between 23 and 38 kilowatts in operation. The iPhone runs on a battery, and charging it from a USB connector can draw a max of 2.5 watts.

More to come…

Tags: CO2, Development, electricity, Energy, iPhone, natural gas, Web