Dr. Jia Di works in the field of digital integrated circuit design, with an interest in using asynchronous logic to make chips run more efficiently.
Most digital circuits are designed with synchronous logic, which means that clocks are used to control the behavior of the chip. Each task is budgeted a certain amount of time, and when that time is up, the next task is sent, regardless of whether the first task was finished. Asynchronous logic, on the other hand, is not controlled by a clock. Instead, each part of the circuit uses what are called handshaking protocols to indicate when it has finished one task and is ready for the next.
"A lot more work has been done on synchronous logic," says Dr. Di, "but actually, asynchronous circuits have some advantages, including low power."
One of Dr. Di's many areas of interest is ultra low-power circuit design. As users of consumer electronics are surely aware, most mobile devices, such as phones, cameras, or computers, spend the majority of their time idle, in some kind of "sleep mode," and nothing is more irritating than discovering that a phone one has not used all day is out of batteries. Ultra low-power circuit design focuses on ways to make these sleep modes even more efficient, to minimize battery drain by devices not in use. So far, his asynchronous circuits have performed more efficiently than the equivalent synchronous circuits, "and that's something we want to explore," he says.
Dr. Di is also using asynchronous logic to help NASA tackle a tough problem for space-bound circuitry: temperature changes. When electronic circuits are exposed to outer space where temperature varies dramatically, perhaps from +180ºC to -230ºC, the operating speed changes significantly. With synchronous logic, this change in speed can cause the chip to fail, because the clock will continue to send tasks at the same rate even though the components of the circuit cannot finish them as quickly. Each rocket can have hundred or thousands of sensors with corresponding data processing elements, each of which could be at risk of failure. Current designs involve housing each piece of electronics in a warm box the size of a small shoebox to keep the temperature inside constant, but each of these boxes is quite heavy, and in space travel, mass adds up very quickly.
However, Dr. Di is hopeful that soon those heavy boxes will be replaced with just a few little chips. An asynchronous digital chip, when exposed to extreme temperature, will exhibit speed changes just as a synchronous one would, but because its operations are not being determined by clocks, the functionality of the chip will not be harmed. In their most recent chip testing, the asynchronous chip worked perfectly from room temperature all the way to two Kelvin, which is -271ºC, without any special adjustment. Another sensor control and data processing chip is currently being manufactured, and soon they will be able to test their final product and, if all goes well, conclude the University of Arkansas' part of the four-year project.
Dr. Di's work in asynchronous logic, low-power electronics, and hardware security all makes for a lot of opportunities for students. "All my projects are done by students," he says. "I have thirteen or fourteen graduate students, and they're all outstanding students and very motivated. I'm really proud of them."