Mohammad Zaman is a DC Consulting Engineer at Nokia with over a decade of experience in IP networking, data center architecture, and network programmability. He specializes in designing scalable, high-performance AI data center fabrics and network architectures that support large-scale GPU clusters, distributed AI training, inference, and high-performance computing (HPC) workloads. His technical expertise spans optimizing GPU and Collective Communications, RCCL/NCCL, BGP EVPN, Segment Routing, RoCEv2 and advanced congestion management mechanisms, including ECN and PFC.
His work focuses on designing, validating, and benchmarking AI/HPC infrastructure, with hands-on experience evaluating GPU cluster performance using distributed AI training, inference, and scientific computing workloads. He is particularly passionate about correlating network behavior with AI/HPC application performance, optimizing fabric efficiency, and leveraging network automation, programmability, and observability to build resilient, and operationally efficient data centers.
Biography
