About the role#
As an AI Performance Engineering Intern, you will join our team to evaluate and optimize AI system performance. You will contribute to the design of test plans and execute experiments to characterize system behavior. This role requires a balance of technical execution, collaboration, and clear communication to ensure project progress.
What you'll need#
- Current enrollment in a B.S. program for Computer Science, Applied Mathematics, Software Engineering, or a related field.
- Coursework or experience with GPU hardware architectures (NVIDIA/AMD) and parallel programming techniques.
- Knowledge of parallel programming methods like NCCL/RCCL, NVSHMEM, or OpenSHMEM.
- Proficiency in Python and familiarity with Linux operating systems, scripting, and performance monitoring tools.
- Background in AI applications and programming models such as PyTorch or TensorFlow.
- Understanding of LLM parallelization methods including TP/PP/DP and EP/MoE.
- Experience with GPU profiling tools like NVIDIA Nsight or AMD rocprof, and familiarity with CUDA/ROCm.
- Experience with AI clusters, job schedulers (PBS, Slurm), and containerization (Docker/Apptainer).
- Familiarity with inference serving engines like vLLM or Ollama and industry benchmarks such as MLPerf.
- Strong analytical, problem-solving, and organizational skills with the ability to multitask.
- Excellent written and verbal communication skills with mastery of the English language.
What you'll do#
- Design and create test plans to evaluate AI system performance.
- Execute experiments to understand the performance characteristics of systems under test.
- Use profiling and monitoring tools to characterize system performance.
- Collaborate with team members to report on progress, share status updates, and resolve issues.
Location & details#
- Location: Spring, Texas, United States
- Modality: On-site
- Employment type: Full-time
- Term: Rolling


