Summary
Tongping Liu is a Principal Engineer and academic with 13 years of experience specializing in runtime systems, OS and compiler techniques to boost performance and safety for multicore and CPU/GPU workloads. He has a strong track record across industry and academia—building non-intrusive GPU memory-management and profiling tools that cut memory use by up to 40% and pinpoint OOM causes, while also publishing novel profilers and deadlock/failure-diagnosis tools. At XPENG he focuses on ML training and inference stability and performance, while maintaining a faculty role at UMass Amherst where his research produced systems for NUMA memory management, cache and synchronization analysis, and automated failure diagnosis. Tongping blends deep systems research with production engineering, delivering solutions that improve throughput for hybrid CPU/GPU training and make hard-to-debug ML failures actionable down to code lines. Notably, many of his tools emphasize zero-code-change deployment, making advanced memory and performance optimizations practical for large ML stacks.
13 years of coding experience
8 years of employment as a software developer
Doctor of Philosophy (Ph.D.) Computer Science, Doctor of Philosophy (Ph.D.) Computer Science at University of Massachusetts Amherst
BA Automatic Measurement and Control, BA Automatic Measurement and Control at Harbin Institute of Technology
Master of Engineering (M.Eng.) Electronics and Information Engineering, Master of Engineering (M.Eng.) Electronics and Information Engineering at Huazhong University of Science and Technology