Quentin Anthony

Quentin Anthony is VP of AI Engineering at Zyphra, where he leads performance optimization for training and inference of novel model architectures on heterogeneous hardware. Quentin has contributed to many open-source model training pipelines, including GPT-NeoX-20B, Pythia, RWKV, RedPajama-INCITE, and Covenant-72B. With over seven years of experience in high-performance computing and distributed deep learning—spanning Ohio State's Network-Based Computing Laboratory, EleutherAI, StabilityAI, and Microsoft Research's DeepSpeed team—Quentin has contributed to the development of hardware-aware co-design of model architectures, GPU kernels, and communication runtimes for higher training throughput and efficiency. He was lead author of the systems technical report for ZAYA1, the first large-scale mixture-of-experts model trained entirely on AMD Instinct MI300X GPUs, AMD Pensando networking, and ROCm open software.