Dr. He is currently a Research Fellow working with Prof. Trevor Mudge in University of Michigan, Ann Arbor. He is participating in research to develop next-generation Software Defined computer architecture which has both high computational efficiency and flexibility to adapt to diverse real-world applications including, but is not limited to deep neural networks and graph processing. He is also conducting research on processor-in-memory accelerated sparse linear algebra, near data processing of graph workloads and in-place fast data format conversion. He also supervises several Ph.D. candidates to explore the hardware infrastructure for autonomous driving, efficient virtual memory design for large scale data center workloads.