Hardware and Software that Allow Computer Servers to Use Renewable Energy More Efficiently


The University of Florida is seeking companies interested in commercializing hardware called \"SolarCore\"" and software called \""iSwitch\"" that will save money by optimizing computer server clusters\' renewable energy use. In the U.S interest has grown in harnessing energy from environmentally sustainable sources such as sunlight for \""greener\"" computing instead of using electricity generated by burning fossil fuels. Photovoltaic systems are expected to provide 15 to 20 percent of the world\'s electricity by 2040. Unpredictable weather events such as cloud cover hamper the systems\' efficiency by blocking sunlight from reaching solar collection panels. This makes it difficult to match power consumption with supply. Available technology incurs significant solar-energy losses since it is largely workload-driven and unaware of supply-side attributes. New harvesting techniques improve energy utilization but they require large expensive batteries. Researchers at the University of Florida have developed hardware and software that significantly improve computer server energy efficiency and performance without the need for batteries. The SolarCore hardware enables multi-core processors to autonomously and efficiently harvest solar energy by performing load matching under variable photovoltaic power output. In laboratory experiments SolarCore achieved an average high green energy utilization score of 82 percent. The iSwitch software allows server clusters consisting of thousands of nodes to switch back and forth from renewable power to traditional utility power (or some other energy source) ensuring that power availability keeps pace with demand. Technology: University of Florida researchers have developed hardware and software for increasing the energy efficiency of renewable-powered server clusters. SolarCore uses multi-core microprocessor power management schemes to achieve high green energy utilization of 82 percent from solar panels under various environmental conditions. Some embodiments include efficient heuristics for allocating the time solar power is used across multiple cores improving workload performance. Joint optimization techniques allow SolarCore to reduce the carbon footprint of computing systems. iSwitch is an energy control system that addresses the operational overhead of renewable energy powered data centers. It leverages existing infrastructures and applies a supply/load cooperative scheme to minimize performance overhead. Together SolarCore and iSwitch optimize renewable energy utilization while effectively managing network traffic and reducing job waiting times. Application: Hardware and software for optimizing microprocessor performance in server clusters that run on renewable energy."


1) Facilitates \"green\"" computing resulting in lower energy costs 2) Performs load matching to variable photovoltaic power output maximizing total solar energy utilization 3) Allows servers to switch from solar (or other intermittent renewable energy sources) to utility power (and back again) when necessary ensuring consistent high-level performance 4) Eliminates the need for batteries enhancing reliability 5) Reduces power dependence on fossil-fuels while optimizing workload performance resulting in a smaller carbon footprint"

Date of release