All jobs
ML Infrastructure Engineer
Nebius
Prague +1 · Remote possibleSenior (5–10 years)Full-timePosted 15 days ago
About the job
What you'll do
Lead and support the benchmarking of GPU platforms to evaluate performance for machine learning and AI workloads. This includes profiling GPU performance at the system and kernel level and optimizing ML workloads to resolve bottlenecks.
What they're looking for
Requires a profound understanding of machine learning theoretical foundations and deep experience with modern frameworks like PyTorch and JAX. Candidates must have strong knowledge of the GPU stack, including CUDA and NCCL, and familiarity with containerized environments.
Full description