Greentech Lead America: Rutgers’s School of Engineering
Building has installed 16 polycrystalline solar panels on top its new computing
facility for research.
The solar panels, known as Parasol, will be used to study
how to effectively manage computing and data processing with solar-generated
electricity.
The project is led by Ricardo Bianchini and Thu D.
Nguyen, professors of computer science. Bianchini has received more than $1.5
million in National Science Foundation grants for research into energy
efficiency in computing.
Further, a team led by Bianchini and computer scientists
at the University of California at Santa Barbara, the University of Michigan,
and the University of Virginia, also received $1.5 million grant from Google
for slashing energy usage in large Internet data centers.
Bianchini and Nguyen are now extending their work from
energy efficiency to renewable energy. Their most recent NSF grant ($420,000),
for “Scheduling Energy Consumption in Green Datacenters,” was awarded last
year.
The research will examine issues surrounding solar
power and computing, such as scheduling computations to maximize use of green
energy by factoring in the amount of daylight hours,
cloud cover, and seasonal changes in the sun’s position. Parasol will be the
platform for their work, and simulations have demonstrated the system’s
feasibility even before it has gone on line, Parashar said.
“Because data centers consume an enormous amount of
electrical power – more than most people realize – making them more
energy-efficient is a great opportunity for us to help reduce the carbon
footprint of computing,” Bianchini said. “And incorporating green power into
the equation, naturally, would be a huge plus.”
Parasol’s rooftop location not only puts the equipment
right under the solar panels, it permits “free cooling,” which is cooling that
does not require energy-intensive compressors or chillers. The system monitors
temperature and energy consumption of all components, and turns off devices
that are not in use. Parasol is wired into the grid and will use “brown energy”
when solar is not available and its batteries are exhausted. The researchers
have developed GreenHadoop and GreenSlot, which are load-scheduling software
systems that maximize the use of green energy.
editor@greentechlead.com