Teddy Do

Teddy Do is a performance engineer on the Transformer Engine team within the NVIDIA DLFW organization. She focuses on accelerating large-scale Transformer training and inference for core frameworks like Megatron-LM, as well as external customers’ models. She joined NVIDIA in 2022 after earning her BS/MS in Computer Engineering from Drexel University, previously working on NVIDIA data center product diagnostics where she designed maximum-stress workloads for single-GPU and rack-scale hardware validation.