UVA research could help make computing more energy-efficient

While the amount of energy needed for computing continues to grow, in just eight years up to 20% of the world’s energy could be consumed by data centers, wireless networks, large electronics audience and a growing number of other devices.

This prediction from Enerdata – which puts current computing power consumption at 9% of global power – is one of the reasons why researchers from the School of Engineering and Applied Sciences at the University of Virginia strive to make computing more energy efficient. This projected growth, coupled with power grids already strained by weather events and the economy’s transition from fossil fuels to renewables, has engineers desperately trying to flatten the energy demand curve of the world. computer science.

The members of Jon Ihlefeld’s multifunctional thin film band are doing their part. They are investigating a materials system that will allow the semiconductor industry to put both memory and computation on a single chip.

“Right now we have a computer chip that is doing its computing activities with some memory on it,” said Ihlefeld, an associate professor of materials science and engineering and electrical and computer engineering.

Every time the computer chip wants to talk to the larger memory bank, it sends a signal down the line, and that requires power. The longer the distance, the more energy is needed. Today, the distance can be up to several centimeters, which in the calculation of distances is really far.

Sherry J. Basler