Sergei Bastrakov
GPU Cluster Architect at Nebius
Baden-Baden, Baden-Württemberg, Germany
Summary
👤
Senior🎓
Top SchoolSergei Bastrakov is a GPU Cluster Architect and CUDA developer with over a decade of hands-on experience building and benchmarking large-scale HPC and AI clusters. He has led technical teams and acceptance performance testing for thousands of H100/H200 GPUs, designed SLURM deployments, NCCL/HPL workflows, and custom health- and monitoring tooling for multi-rail InfiniBand systems. A former computational scientist and core developer of the exascale PIConGPU particle-in-cell code, he blends deep low-level C++/GPU expertise with practical cloud delivery and vendor collaboration. His work spans root-cause performance tuning (NUMA, BIOS, core pinning, GPUDirect RDMA) and productionizing NVIDIA AI Enterprise-based products and PoCs. Based in Baden-Baden, Germany, he pairs a PhD-level research background with real-world cloud-scale GPU operational leadership, and has contributed notable performance-portable improvements to the widely used picongpu project.
11 years of coding experience
12 years of employment as a software developer
Doctor of Philosophy - PhD Computer Science, Doctor of Philosophy - PhD Computer Science at State University of Nizhni Novgorod named after N.I. Lobachevsky (UNN)
Russian, English, German
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (9)
c-language10
cprogramming-language10
template-metaprogramming10
message-passing8
mpi8
cuda8
numerical-methods8
parallel-computing7
performance-optimization7
Programming languages (5)
Github contributions (5)
ComputationalRadiationPhysics/picongpu
Jan 2018 - Sep 2022
Performance-Portable Particle-in-Cell Simulations for the Exascale Era :sparkles:
Role in this project:
Backend Developer
Backend DeveloperContributions:976 reviews, 878 commits, 594 PRs in 4 years 8 months
Contributions summary:Sergei primarily contributed to the PMacc framework within the repository, adding new functionalities and improving existing ones. They added a new file providing the value of pi, fixed template parameter usage in a buffer cursor, and addressed formatting issues within the code. They also refactored parts of the codebase, including implementations related to particle boundaries and the handling of current deposition.
laserparticle-in-cellsimulationpicexascale
sbastrakov/picongpu
Jan 2018 - Sep 2022
A particle-in-cell code for GPGPUs :sparkles:
Contributions:602 pushes, 673 branches in 4 years 8 months
particleparticle-in-cellsparkles
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.