As a child, I was always curious about the world around me; I wanted to be a scientist when I grew up. I liked disassembling cassette players, radios, and TVs to figure out why and how they worked, and then I tried to put them back together (not always successfully). However, my parents always supported me, bought me tools to disassemble our home electronics, and told me inspiring stories about the greatest scientists in the world. Through my life, I was encouraged to be whoever I wanted to be.
After I received a B.S. degree in microelectronics from Peking University, Beijing and a Ph.D. degree in electrical engineering from the Pennsylvania State University in 2014, I joined Intel Labs to work in process-optimized microarchitecture design. I’m now a staff research scientist in the Intel Artificial Intelligence Products Group, exploring circuits and architecture interactions that enable emerging logic and memory technology for energy-efficient applications and agile hardware design for machine learning accelerators.
On Monday, June 3, I’ll be one of five recipients of an Under-40 Innovators Award, sponsored by the Association for Computing Machinery (ACM) and the Institute of Electrical and Electronics Engineers (IEEE) at the Design Automation Conference in Louisville, CO. I’m being honored for my work adapting magnetoelectric spin orbit (MESO) technology for AI.
MESO is fundamentally different from the complementary metal–oxide–semiconductor (CMOS) technology we’ve been using in processors and memories for more than 50 years. MESO uses electronic spin instead of electrical charge to store information, allowing devices to be 10 to 30 times more energy efficient and five times smaller than the limits of CMOS. Even though the basic physics of MESO is beyond CMOS, MESO is much more conventional than quantum computing technology; it adapts to traditional computer architecture and operates at room temperature.
At Intel, I discovered ways to use MESO technology in a complete set of logic and non-volatile memory devices which can be integrated on one chip, or theoretically, with CMOS circuits in a single device. I contributed to an article, “Scalable Energy-efficient Magnetoelectric Spin–orbit Logic” published in Nature in December 2018 which referenced my circuit designs, modeling and simulation.
Thanks to Intel’s diversity in the technology business, my job provides me with incredible opportunities to innovate and transform ideas into reality. Working for Intel gives me resources to develop the state-of-art technology in computing, collaborate with experts in different fields and enable future solutions to real-world problems. The work is full of challenges, but I’m proud to have realized my dreams of becoming a scientist – and a driving force in technology innovation.