CUDA Core Libraries Intern

Job Description

CUDA Core Libraries Intern | NVIDIA

The Tone:
This is an internship at NVIDIA, a company dedicated to visual and AI computing, pioneering the GPU to solve complex problems in AI, deep learning, robotics, and self-driving cars. This role contributes to the foundational CUDA Core Libraries, which are C++ and Python libraries enabling developers to write fast, reliable, and scalable GPU-accelerated software for modern HPC and AI. The intern’s work directly impacts the build, testing, packaging, and developer experience, accelerating development and delivery of these critical components for a wide range of workloads.

The TL;DR
• Role: Internship
• Pay: $20–$71 hourly
• Team: The CUDA Core Libraries team, which builds, tests, and packages foundational libraries, algorithms, language, and compiler infrastructure.
• Mission: This intern will contribute to the build, testing, packaging, and developer experience to accelerate development and deliver CUDA Core Libraries for both C++ and Python developers.
• Tech Stack: CMake, scikit-build-core, Conda, PyPi, GitHub, GitLab, Docker, C++, CUDA, Python, Thrust, CUB, libcudacxx, cuda-python, numba-cuda

What You’ll Actually Do
• Build Processes: Decompose and modularize build processes for reusability across multiple projects.
• Deployment: Partner with engineering teams to ensure their code can be built and deployed across Conda and PyPi ecosystems on Linux and Windows.
• Packaging: Develop robust and modern approaches to packaging compiled code, Python wrappers, and their dependencies for user distribution.
• CI Pipelines: Design CI pipelines that enable rapid build and testing of new code to improve development velocity and intelligently sample architecture, OS, and GPU coverage.
• Collaboration: Collaborate with expert CUDA engineers, participating in design reviews, code reviews, and open-source-style workflows.

The Must-Haves
• Background: Currently pursuing a BS, MS, or PhD in Computer Science, Computer Engineering, or a related field.
• Experience: Experience with build systems such as CMake or scikit-build-core, along with packaging in Conda and/or PyPi; familiarity with CI/CD systems including GitHub, GitLab, or other related platforms and the use of Docker images to facilitate workflows.
• Skills: Familiarity with debugging the output of build systems and compilers to resolve issues in complex build environments involving modern C++, CUDA, and/or Python libraries; experience with software libraries or open-source projects, including testing, performance profiling, and code reviews; ability to work independently and drive a project from exploration to completion; clear written communication for design discussions and documentation.
• Bonus: Knowledge of CPU/GPU architecture and how hardware details impact algorithmic performance; familiarity with binary library compilation, linking, packaging, distribution, ABI compatibility, and deployment strategies on Linux and/or Windows; familiarity with compiler infrastructure and tooling such as LLVM, Clang/LLVM tooling, or MLIR; comfort navigating and debugging large, multi-language codebases (C++, Python, CMake, GitHub Actions CI systems); demonstrated interest in developer tools, library design, developer experience, and making other developers faster and more productive.