Author

Xiaomin Lu is currently a Principal Member of Technical Staff at AMD in Austin TX. His primary areas include GPU performance architecture, workload performance profiling and bottleneck analysis methodology. He architected and led the development of the MIPerf profiling tool for MI100/MI200 GPUs. Prior to joining AMD, he was a performance architect at NXP Semiconductors working on math, machine learning and packet processing accelerators in Layerscape and S32 family silicons deployed in the automotive and networking industry. Outside of work he enjoys cooking and road cycling in the Texas Hill Country areas.