Sreeram Potluri
Principal Software Architect at NVIDIA
California, United States
Summary
🤩
Rockstar🎓
Top SchoolSreeram Potluri is a Principal Software Architect at NVIDIA with nine years of focused experience building low-latency communication runtimes and GPU-cluster technologies for HPC, deep learning, and packet processing. He has led teams responsible for GPUDirect and libraries such as NCCL, UCX/OpenMPI, and NVSHMEM—a PGAS runtime he helped design to enable fine-grained, in-kernel GPU communication and bypass the CPU. His work bridges academic research (Ph.D. work at Ohio State on MPI/PGAS and MVAPICH2/MVAPICH2-X) with production-grade systems that power large InfiniBand and heterogeneous GPU clusters. Based in California, he combines deep protocol and interconnect expertise with hands-on product delivery, and is known for enabling direct GPU control of network devices (e.g., GPUDirect Async/GPU Verbs)—a less visible but critical enabler of modern GPU-native networking.
9 years of coding experience
16 years of employment as a software developer
Jawaharlal Nehru Technological University HyderabadDoctor of Philosophy (Ph.D.) Computer Science and Engineering, Doctor of Philosophy (Ph.D.) Computer Science and Engineering at The Ohio State University
Github Skills (32)
mpi10
cluster-computing10
nvidia10
multi-gpu10
cuda10
testing9
infiniband9
networking9
4g9
gpu9
mailman9
yang9
shared-memory9
portability8
ospf8
Programming languages (4)
Github contributions (5)
spotluri/libmp
Feb 2020 - Jan 2021
Simple message passing library
Contributions:1 PR, 15 pushes, 1 branch in 1 year
messagemessage-passingsimple-messagepassing
spotluri/openshmem-specification
Feb 2017 - Nov 2017
Contributions:6 PRs, 18 pushes, 1 branch in 9 months
openshmem
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.