Donghua Jiang is a Senior Inference Engineer on Tencent’s Hunyuan team and holds a master’s degree from Peking University. His work focuses on high-performance computing, inference-system optimization, and GPU kernel development. As a core developer, he has contributed to inference optimization across multiple generations of Hunyuan models and to the development and open-sourcing of HPC-Ops, a high-performance kernel library used as an official backend by vLLM and SGLang.