HPC Software & Ops Intern: Learn Multi-Node GPU

InternHunt; Lausanne; Hybrid; CHF 13,000 - 20,000; Full time

Job summary

Corintis in Lausanne offers an HPC Software & Operations Intern position focused on operating, optimizing and developing our HPC environment across multi-core, multi-GPU, multi-node systems. This hands-on role starts with assisted work and grows toward autonomy, with mentorship from infrastructure and software teams.

You will learn MPI tuning, workload decomposition, and containerised HPC environments while contributing to runbooks and performance analysis.

Qualifications

• Knowledge of Linux and of Python, Bash, C or C++.
• Knowledge of, or interest in, MPI and parallel or distributed computing.
• Knowledge of, or interest in, GPU computing (CUDA).
• Strong problem-solving skills.

Responsibilities

• Adapt existing code and engineering workflows to the HPC environment, including multi-core, multi-GPU and multi-node execution.
• Write and maintain launch scripts and job scripts, and work with the scheduler and the job queues.
• Run and troubleshoot MPI workloads, and analyse jobs that fail or perform poorly.
• Benchmark workloads, test their scalability and validate them across multiple nodes.
• Analyse CPU, GPU, memory, network and I/O utilisation to find bottlenecks and scaling limits.
• Manage HPC software environments, modules and dependencies.
• Support cluster users with their workloads and with data staging, data access and storage workflows.
• Take part in day-to-day cluster operations and incident analysis, together with the infrastructure and software teams.
• Write procedures, best practices and runbooks.
• Analyse failed or poorly performing jobs
• Work on MPI tuning, workload decomposition, CPU/GPU affinity, UCX, profiling, multi-node GPU execution and containerised HPC environments.

Skills

Linux

Python

Bash

C/C++

MPI

CUDA

Parallel computing

Problem-solving

Education

Student in Computer Science or related field

Tools

MPI

CUDA

Job description

Corintis in Lausanne offers an HPC Software & Operations Intern position focused on operating, optimizing and developing our HPC environment across multi-core, multi-GPU, multi-node systems. This hands-on role starts with assisted work and grows toward autonomy, with mentorship from infrastructure and software teams.

You will learn MPI tuning, workload decomposition, and containerised HPC environments while contributing to runbooks and performance analysis.