Rocking Lai is a Machine Learning Software Engineer with nine years of hands-on experience building high-performance computer vision and deep learning systems across GPU, NPU/TPU and embedded platforms. Currently at AMD, he develops and optimizes GPU backends and complex fused kernels for frameworks like PyTorch and TensorFlow, with a strong emphasis on low-precision arithmetic and quantization. Previously he designed quantization toolchains, mix-precision algorithms (patented) and TPU simulators at Bitmain, and built full-stack video surveillance and vision systems earlier in his career. He contributes to performance-critical open-source work—enhancing GEMM shuffle kernels and int8/bf16/fp32 support in the ROCm composable_kernel project—demonstrating a blend of low-level kernel engineering and system design. Comfortable collaborating with IC designers and leading compiler frontends, he pairs research-driven prototyping with production-grade implementation. Based in New Taipei, Taiwan, he also mentors and shares knowledge in computer vision and ML education.
9 years of coding experience
7 years of employment as a software developer
Bachelor's degree, Computer Science, Bachelor's degree, Computer Science at 國立交通大學
Composable Kernel: Performance Portable Programming Model for Machine Learning Tensor Operators
Role in this project:
Back-end Developer
Contributions:497 reviews, 356 commits, 115 PRs in 1 year
Contributions summary:Rocking primarily contributed to the implementation and enhancement of the `DeviceGemmXdl_C_Shuffle` and related functionalities within the `rocm/composable_kernel` repository. Their work involved adding and refining host APIs, including those for shuffle versions of DeviceGemmXdl. They implemented support for alpha/beta scaling and integrated new features to existing functionalities. The user also worked on supporting and refining the int8, bf16, and fp32 gemm instances.
Contributions:181 commits, 4 PRs, 195 pushes in 6 months
streamingcross-platformelectronplayer
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.
Request Free Trial
Rocking Lai - Machine Learning Software Engineer at AMD