Release Date: June 5, 2013
BUFFALO, N.Y. – Jonathan Bird and Uttam Singisetti don’t act like renegades – both dress neatly and are polite in conversation.
But steer the discussion to computing power, specifically society’s uncompromising demand for more of it, and the radical in both University at Buffalo researchers begins to surface.
“Industry’s ability to build faster, less expensive and more powerful computers will likely reach an end in 10 or 15 years” said Bird, UB professor of electrical engineering. “The reason: electrical transistors that make computing possible are getting so small they’ll soon be on an atomic level. There’s nowhere to go from there.”
That is, unless you change how electronics work.
Bird and Singisetti, UB assistant professor of electrical engineering, are the recipients of a five-year, $850,000 grant that aims to do just that. Part of a larger $7.1 million effort based at the University of Nebraska-Lincoln’s Center for NanoFerroic Devices (CFND), the UB grant will fund low-power computing experiments that involve magnetoelectric and ferroelectric devices.
Before describing the research, here’s a primer on modern electronics and spintronics, a promising field of study that relates to the work of Bird and Singisetti.
Modern electronic devices record and read data via negative and positive electrical charges that are represented, in circuits, as ones or zeros. Processing information requires moving electrons, which consumes energy and produces heat.
Electrons have another property that responds to magnetic fields. This property, known as spin, creates a tiny magnetic field. The spin is characterized by an “up” or “down” direction which, like the electrical charges, can be used to encode data. This is the basis of spintronics.
Magnetoelectric devices use the spin property of electrons; however, researchers are going a step further by manipulating the spin with materials such as chromium oxide. They then apply an electric field that changes the spin state without moving electrons, thus reducing the energy of the operation.
Unlike magnetoelectric devices, ferroelectric devices do not use spin or magnetism. Instead, they are created by applying an electric field that reverses their naturally-occurring charge. The process is similar to how common magnets, like those used on refrigerators, work.
Both types of devices are advantageous, Bird and Singisetti said, because neither device requires the movement of electrons nor will data disappear when there is no power applied. In theory, each could lead to faster and more powerful computing while consuming less energy.
“Just as vacuum tubes led to transistors, transistors may give way to magnetoelectric and ferroelectric devices creating a whole new class of computers and other devices capable of executing tasks that we can only dream about today. The implications are huge,” Singisetti said.
Making advancements in the field is a priority of big businesses like GlobalFoundries, IBM, Intel, Micron Technology and Texas Instruments - all of which support a public-private partnership called the Nanoelectronics Research Initiative (NRI).
The NRI was created by the university-research consortium Semiconductor Research Corp. and the National Institute of Standards and Technology, an agency of the U.S. Department of Commerce that promotes innovation and industrial competitiveness.
The NRI will spend more than $23 million over nearly five years at three multi-university research centers including the CFND. The two other multi-university research centers are based at the University at Albany and The University of Texas at Austin.
In addition to UB and the University of Nebraska-Lincoln, there are four other universities working with the CFND on this project. They are: the University of Wisconsin-Madison, Oakland University, the University of California-Irvine and the University of Delaware.
Bird and Singisetti will be fabricating nanosized devices in the Davis Hall clean room and testing their electrical characteristics. They intend to work with UB’s Center of Excellence in Materials Informatics, which was established in 2012 with funding from New York State.
In addition to supporting their research, the $850,000 will fund two PhD students each year.
Bird and Singisetti are members of the UB Department of Electrical Engineering’s Solid State Electronics Research Group, which conducts research in nanoelectronics, nanomaterials and characterization, terahertz technology, microelectronic and optoelectronic devices and materials, carbon nanotubes, plasma dynamics, transport and device physics in semiconductor heterostructures, magnetism, very-large-scale integration (VLSI) and field-programmable gate array (FPGA) circuits, and systems-on-a-chip.
Other members of the group include Liesl Folks, dean of the UB School of Engineering and Applied Sciences; professors Ping-Chin Cheng, Vladimir Mitin, David Shaw, Chu Ryang Wie; associate professor Kwang Oh; and assistant professor Cristinel Ababei.