Summary
Xin Chen is a machine learning software engineer and systems architect with over a decade of hands-on experience building large-scale AI, computer vision, and efficient on-device models. Currently at Intel, he previously led research and engineering across industry and academia—designing petascale training platforms, lightweight compiler-aware model optimizations, and TVM-based tensor program tuning at companies including Midea, Kwai, and Petuum. He has a strong background in heterogeneous computing (CPU/ARM/NVIDIA CUDA), GPU cluster design, and realistic productionization of vision models for mobile and edge. His work spans applied research (face landmarks, GANs, temporal grounding) to systems-level contributions like dynamic pruning, distillation, and a cost-efficient 128-GPU training rack referenced in patents and papers. Mentoring interns from top universities and publishing in top venues, he blends deep academic training (PhD) with pragmatic engineering that squeezes performance out of constrained hardware. Based in California, he quietly bridges research and deployment—turning algorithmic ideas into optimized, production-grade systems.
12 years of coding experience
11 years of employment as a software developer
Master, Electrical Engineering(Intelligence Control and Pattern Recognition), Master, Electrical Engineering(Intelligence Control and Pattern Recognition) at Hefei University of Technology
Ph.D, Ph.D at University of Hawaii at Manoa