Shahnawaz Ahmed is a Senior Deep Learning Researcher with a focus on making large neural networks run efficiently on edge hardware. His work encompasses quantization, pruning, neural architecture search, and inference optimization, enabling models developed in PyTorch to achieve real-time execution on devices such as NVIDIA GPUs (Orin, Thor), AMD Strix Halo, and Qualcomm NPUs.
Ahmed is actively involved in optimizing vision-language-action models, such as Physical Intelligence π-0.