Experiments show magnetic chips could dramatically increase computing’s energy efficiency

Beyond Moore’s law: the challenge in computing today is reducing chips’ energy consumption, not increasing packing density
March 17, 2016

Magnetic microscope image of three nanomagnetic computer bits. Each bit is a tiny bar magnet only 90 nanometers long. The image hows a bright spot at the “North” end and a dark spot at the “South” end of the magnet. The “H” arrow shows the direction of magnetic field applied to switch the direction of the magnets. (credit: Jeongmin Hong et al./Science Advances)

UC Berkeley engineers have shown for the first time that magnetic chips can actually operate at the lowest fundamental energy dissipation theoretically possible under the laws of thermodynamics. That means dramatic reductions in power consumption are possible — down to as little as one-millionth the amount of energy per operation used by transistors in modern computers.

The findings were published Mar. 11 an open-access paper in the peer-reviewed journal Science Advances.

This is critical at two ends of the size scale: for mobile devices, which demand powerful processors that can run for a day or more on small, lightweight batteries; and on an industrial scale, as computing increasingly moves into “the cloud,” where the electricity demands of the giant cloud data centers are multiplying, collectively taking an increasing share of the country’s — and world’s — electrical grid.

“The biggest challenge in designing computers and, in fact, all our electronics today is reducing their energy consumption,” aid senior author Jeffrey Bokor, a UC Berkeley professor of electrical engineering and computer sciences and a faculty scientist at the Lawrence Berkeley National Laboratory.

Lowering energy use is a relatively recent shift in focus in chip manufacturing after decades of emphasis on packing greater numbers of increasingly tiny and faster transistors onto chips to keep up with Moore’s law.

“Making transistors go faster was requiring too much energy,” said Bokor, who is also the deputy director the Center for Energy Efficient Electronics Science, a Science and Technology Center at UC Berkeley funded by the National Science Foundation. “The chips were getting so hot they’d just melt.”

So researchers have been turning to alternatives to conventional transistors, which currently rely upon the movement of electrons to switch between 0s and 1s. Partly because of electrical resistance, it takes a fair amount of energy to ensure that the signal between the two 0 and 1 states is clear and reliably distinguishable, and this results in excess heat.

Nanomagnetic computing: how low can you get?

The UC Berkeley team used an innovative technique to measure the tiny amount of energy dissipation that resulted when they flipped a nanomagnetic bit. The researchers used a laser probe to carefully follow the direction that the magnet was pointing as an external magnetic field was used to rotate the magnet from “up” to “down” or vice versa.

They determined that it only took 15 millielectron volts of energy — the equivalent of 3 zeptojoules — to flip a magnetic bit at room temperature, effectively demonstrating the Landauer limit (the lowest limit of energy required for a computer operation). *

This is the first time that a practical memory bit could be manipulated and observed under conditions that would allow the Landauer limit to be reached, the authors said. Bokor and his team published a paper in 2011 that said this could theoretically be done, but it had not been demonstrated until now.

While this paper is a proof of principle, he noted that putting such chips into practical production will take more time. But the authors noted in the paper that “the significance of this result is that today’s computers are far from the fundamental limit and that future dramatic reductions in power consumption are possible.”

The National Science Foundation and the U.S. Department of Energy supported this research.

* The Landauer limit was named after IBM Research Lab’s Rolf Landauer, who in 1961 found that in any computer, each single bit operation must expend an absolute minimum amount of energy. Landauer’s discovery is based on the second law of thermodynamics, which states that as any physical system is transformed, going from a state of higher concentration to lower concentration, it gets increasingly disordered. That loss of order is called entropy, and it comes off as waste heat. Landauer developed a formula to calculate this lowest limit of energy required for a computer operation. The result depends on the temperature of the computer; at room temperature, the limit amounts to about 3 zeptojoules, or one-hundredth the energy given up by a single atom when it emits one photon of light.


Abstract of Experimental test of Landauer’s principle in single-bit operations on nanomagnetic memory bits

Minimizing energy dissipation has emerged as the key challenge in continuing to scale the performance of digital computers. The question of whether there exists a fundamental lower limit to the energy required for digital operations is therefore of great interest. A well-known theoretical result put forward by Landauer states that any irreversible single-bit operation on a physical memory element in contact with a heat bath at a temperature Trequires at least kBT ln(2) of heat be dissipated from the memory into the environment, where kB is the Boltzmann constant. We report an experimental investigation of the intrinsic energy loss of an adiabatic single-bit reset operation using nanoscale magnetic memory bits, by far the most ubiquitous digital storage technology in use today. Through sensitive, high-precision magnetometry measurements, we observed that the amount of dissipated energy in this process is consistent (within 2 SDs of experimental uncertainty) with the Landauer limit. This result reinforces the connection between “information thermodynamics” and physical systems and also provides a foundation for the development of practical information processing technologies that approach the fundamental limit of energy dissipation. The significance of the result includes insightful direction for future development of information technology.