Lorri Rao

Lorri Rao specializes in LLM training and inference optimization, focusing on GPU kernel and network communication optimizations. Prior to AMD, Lorri got PhD degree in computer architecture at USC (2020-2024), and BS at HUST (2016-2020).

Posts by Lorri Rao

September 17, 2026

Implementing a High-Performance Custom Diffusion Attention Kernel with FlyDSL

Learn how to implement and optimize flexible, high-performance diffusion attention kernels with FlyDSL.

https://rocm.blogs.amd.com/software-tools-optimization/flydsl-customized-attention/README.html

December 02, 2025

Optimizing LLM Workloads: AMD Instinct MI355X GPUs Drive Competitive Performance

Explore ROCm 7.0’s AI training boost! See how MI355X accelerates JAX and PyTorch frameworks to unlock faster and efficient LLM scaling.

https://rocm.blogs.amd.com/artificial-intelligence/ROCm7-MI355X-training-performance/README.html