Senior Formal Verification Engineer, GPU Kernels

Details

Modern AI performance relies on highly optimized GPU kernels—performance-critical code where bugs can be hard to catch and expensive to miss. NVIDIA's Deep Learning Safety Team is hiring engineers to build verification tools that prove these kernels behave correctly, enabling their deployment in a large range of environments, including safety-critical systems. The mission is to design and develop scalable verification tools for GPU kernels that can handle the massive concurrency and complex memory model of the latest GPU architectures.

Formal methods alone cannot scale to modern GPU kernels, and AI alone cannot offer safety guarantees. The team combines both approaches to support compiler and kernel developers working on safe autonomous driving.

Responsibilities

• Design and develop robust and scalable verification tools for GPU kernels.
• Integrate verification work into production pipelines to support kernel and compiler developers.
• Integrate AI into formal verification workflows.
• Build agents to automate verification tasks, including formalization of specifications, bug fixing, and root cause analysis.
• Develop new verification algorithms and evaluate them.
• Build tools to automate workflows.
• Participate in architecture discussions.
• Read research papers, prototype ideas, and contribute to the research community where appropriate.
• Practice hardware-software co-design while developing innovative software and hardware products.

Requirements

• MS or PhD in Computer Science, Computer Engineering, or equivalent experience.
• 6+ years of relevant work experience.
• Experience with formal methods, such as symbolic execution, SMT solving, interactive theorem proving, or model checking.
• Strong programming skills in C/C++ or Rust.
• Experience with source code management systems such as Git.
• Experience with build systems such as Make or CMake.
• Ability to work independently, define project goals and scope, and lead an individual development effort.

Preferred Qualifications

• Knowledge of CPU and/or GPU architecture.
• CUDA or OpenCL experience.
• Background in the formalization of weak memory models.
• Experience verifying concurrent software.
• Experience building LLM agents with tool use and multi-step reasoning.
• Experience with neurosymbolic approaches or LLM-assisted theorem proving.

Benefits

The role includes eligibility for equity and benefits.