AI Compiler Formal Verification Intern

NVIDIA, renowned for inventing the GPU in 1999, has been a driving force in the PC gaming market, computer graphics, and parallel computing. The company is now widely recognized as "the AI computing company," with GPUs serving as the brains behind modern AI applications, including robots and self-driving cars. The AI Compiler Formal Verification Intern will focus on researching and implementing innovative formal methods to verify state-of-the-art AI compilers. This includes working on various critical elements such as compiler features, passes, IRs, sub-graphs, and lowerings. The intern's main responsibilities will involve exploring and applying AI, formal verification, and other semantic verification or static analysis techniques. The goal is to develop solutions that verify AI compilers for both human and agentic-use cases. This role offers a unique opportunity to contribute to cutting-edge technology in a dynamic and innovative environment. To qualify for this role, candidates must be pursuing a PhD in Computer Science, Computer/Electrical Engineering, Mathematics, or a related field. Strong programming skills in Python or C++ are essential, along with a robust academic background in formal methods, preferably at the PhD level. The role is based in a forward-thinking and rapidly growing engineering team, offering competitive salaries and a generous benefits package. Nice-to-have skills include an academic or industrial background using LEAN or SMT to solve compiler or compiler-adjacent problems. Knowledge of deep learning frameworks such as Pytorch and TensorRT, as well as compilers, will also be beneficial. The team at NVIDIA is known for its creativity and passion for technology, making it an exciting place to work and grow. What the role asks for: - Pursuing a PhD in Computer Science, Computer/Electrical Engineering, Mathematics or equivalent program - Strong Python or C++ programming skills - Strong academic background in formal methods (preferably PhD) - Nice-to-have: Academic or industrial background using LEAN or SMT to solve compiler or compiler-adjacent problems - Nice-to-have: Knowledge of deep learning frameworks (such as Pytorch and TensorRT) and compilers