Professor Xue's research focus is on storage and systems, and high-performance system designs for AI. He is one of the pioneer researchers to consider Non-Volatile Memories (NVMs) at the software and system level. NVM, such as Phase Changing Memory (PCM), STT-RAM, and FeRAM, shows huge potential because of its characteristics of high density, shock-resistance, and near-zero leakage power. NVM has shortcomings such as long write latency, high write energy, and limited endurance. This property of asymmetric read/write operations poses unique challenges. Professor Xue's (DAC’10, DAC’11) address read/write asymmetry from the system and software level when NVM is applied in the memory hierarchy. NVM writes are further calibrated with different intensities (LCTES’13, DATE’13), which leads to optimized power consumption, performance, and lifetime. These works have profound impacts on widely applied commercial products such as Intel 3D X-point memory and TI MSP series today.