I recently spoke with representatives at Hive Computing, Inc. – a fledgling start-up with a revolutionary idea that meshes smart meters and computing, and that may take the shared economy to a new level. Hive plans to connect smart meters and utilize their spare capacity to create a networked supercomputer that will be green and very inexpensive to operate.
According to Eric Frazier, Co-Founder of the start-up company, every smart meter essentially has the computing capability of a cellphone. Every few minutes these meters send usage data to the utility. However, 99% of the time they just sit idle. But if you could leverage the idle time of all of these devices, you would have thousands of central processing units with multiple cores available for parallel processing. And the more cores you have, the faster you can process information.
Hive is on its way to doing just that. So in the near future, that little meter sitting outside your house might be buzzing away, solving a problem related to optimization of the smart grid, analyzing an individual’s genome to come up with a personalized cancer treatment regime, or performing many thousands of other tasks that require a supercomputer, but that up until now have been too expensive to perform.
Frazier indicates that 10,000 meters gets you into the range of a supercomputer with a teraflop (1 trillion floating operations per second) of processing power. A million meters represents two petaflops (two quadrillion operations per second) of capacity. Raiford Smith, Vice President, Corporate Development and Planning at CPS Energy in San Antonio, thinks the idea has promise.
"With the advent of the Internet of Things and the Smart Grid, utilities are deploying millions of intelligent devices – devices which, when interconnected, can form the basis of an impressive super computing platform. In fact, a 1 million meter deployment would be the equivalent of the world’s 20th fastest super computer. This represents an opportunity to do something good for society by growing low-cost computing capability while giving utilities an avenue to further invest in their metering infrastructure."
As Frazier notes, the U.S. alone currently has over 50 million meters, which would equal 100 petaflops. To put that in perspective, the world’s largest computer – the Tianhe supercomputer in Guanzhou, China currently boasts 33.8 petaflops of maximal achieved performance (and draws 18 MW of power); the largest U.S. machine the Cray Gemini at Oak Ridge is about half that size.
Frazier observes that this 100 petaflop number will likely increase in the near future as smart meters evolve. “That’s with existing single core processors. Most new meters will have four processors.” That increases the firepower significantly.
In addition to the novel shared resource approach, two aspects of the Hive approach are extremely intriguing:
First, this supercomputer represents an extremely clean resource. According to Angel Orrantia, Co-Founder,
Our computer will be on the Top 500 Green List. We don’t consume any additional power; smart meters are already on 24/7. Moreover, the system does not require cooling. Cooling of existing supercomputers and data centers is the largest use of power. The Hive solution creates supercomputing capability without requiring additional resources or associated carbon. In addition, if you think about all of the Scope 3 type carbon consumption, the logistics to get all of the equipment to the data center and build all of the infrastructure, there is a tremendous amount of carbon footprint…in this scenario we leverage infrastructure already in place.
Second, this will be by far the cheapest supercomputer in the world. A new supercomputer can cost well into the tens of millions of dollars. For example, the U.S. National Oceanic and Atmospheric Administration (NOAA) announced last January an allocation of $44.5 million to increase its computing capacity to five petaflops. Similarly, Britain’s weather forecasting agency is investing 97 million British Pounds Sterling into its massive supercomputer.