Intern - AI Cluster Engineering (Summer 2026)
About The Position
We are building the AI DC level software framework. It is a cutting-edge platform to validate next-generation, full-stack AI infrastructure. We are looking for talented interns to develop the AI application workload framework. Beyond AI models, AI applications will be containerized, optimized, and stress-tested on versatile GPU clusters to find architectural bottlenecks. We are looking for interns who will onboard and optimize diverse AI Applications onto our AI cluster, deep-diving into specific domains like Generative AI (LLM), Physical AI (Robotics), and AI for Science (Bio/Physics) to benchmark their performance on the latest GPUs, DPUs, and Network fabrics.
Requirements
- Education: Currently pursuing a MS, or PhD in Computer Science, Electrical Engineering, AI, or related fields.
- Programming: Strong proficiency in Python (Bash scripting is a plus).
- Containerization: Experience with Docker (building Dockerfiles, managing dependencies).
- AI Fundamentals: Basic understanding of Deep Learning workflows (Training vs. Inference) and frameworks (PyTorch, TensorFlow).
- OS: Comfort with Linux command-line environment
Nice To Haves
- Orchestration: Experience with Kubernetes (K8s) or Slurm.
- Profiling: Experience with performance profiling tools (Nsight Systems, DCGM, PyTorch Profiler).
- Domain Expertise (in one of the following):
- LLM: Experience with vLLM, TGI, or TensorRT-LLM.
- Distributed Systems: Experience with Multi-node training (Megatron-LM, DeepSpeed).
- Robotics: Experience with ROS 2, Isaac Sim, or Reinforcement Learning.
- HPC/Science: Experience with MPI, OpenMP, or Bio-informatics tools.
- Hardware: Curiosity about computer architecture (GPU, DPU, Memory, Network Fabric).
Responsibilities
- Onboard Diverse AI Applications: Port and deploy state-of-the-art AI workloads, including LLM (vLLM, TGI), Physical AI (Isaac Sim, ROS 2), and Scientific AI (AlphaFold, GROMACS).
- Resolve software dependency issues and build optimized Docker images for the A^3 Registry.
- Application Profiling & Analysis: Analyze the unique resource patterns of each application (e.g., “Isaac Sim requires heavy ray-tracing capability” or “Megatron-LM is bottlenecked by RDMA”).
- Identify performance bottlenecks using profiling tools (Nsight Systems, PyTorch Profiler).
- Develop "Test Recipes": Define the standard configuration (Recipe) for each application to ensure reproducible testing.
- Collaboration: Work with infrastructure engineers to tune the system (OS, Network, Storage) to fit your application's needs.
Benefits
- Eligible interns will receive a housing allowance during their internship.
Stand Out From the Crowd
Upload your resume and get instant feedback on how well it matches this job.
What This Job Offers
Career Level
Intern
Number of Employees
5,001-10,000 employees
