Researchers devise extremely energy efficient microprocessor using superconductors

In a new development, researchers at the Yokohama National University have created a prototype of a microprocessor using superconductor devices. Distinctly, energy efficiency of the microprocessor is 80 times more than that of avant-garde semiconductor devices found in microprocessors of currently used high-performance computing systems.

Meanwhile, with increasing integration of avant-garde technologies in daily lives, the need for computational power continues to increase. Therefore, with increasing use of computational systems, the energy needed to run these systems is growing immensely. For example, to serve the energy need of modern day data centers, some of them are constructed near rivers to make use of flowing water to cool the machinery.

In fact, the digital communications infrastructure in place that supports today’s Information Age currently approximately uses 10% of the global electricity. Meanwhile, studies present an alarming scenario pertaining to use of electricity by computing systems by 2050. It is estimated electricity usage to rise more than 50% by 2050 if the underlying technology for communications infrastructure does not advance.

The details of the research published in IEEE Journal of Solid-State Circuits covers the effort to develop an architecture for microprocessors to be energy efficient. Such an architecture of microprocessors would involve use of superconductors, devices that are highly efficient, however, require certain environmental conditions to function.

Hence, to address the power problem, the use of an extremely energy-efficient superconductor digital electronic structure explored. Called adiabatic quantum-flux-parametron, the structure serves as a building block for high-performance, ultra-low-power microprocessors, and other computing systems for next-generation data centers and communication networks.

Through the paper, researchers wanted to prove how Adiabatic quantum-flux parametron enble energy-efficient computing from a practical viewpoint.