Senior GPU Compiler Development Engineer

About the Role

We invite an experienced compiler engineer to join our GPU Assembler team focusing on creating cutting-edge tools that enable highly performant applications and libraries across HPC, deep learning, and autonomous driving sectors. This role involves developing core components accelerating general purpose computation on NVIDIA GPUs and supporting thousands of global CUDA users through the Disassembler (NVDISASM) and Assembler tools designed for optimal performance on GPU architectures.

Key Responsibilities

• Lead the design and implementation of NVIDIA's GPU compute Assembler and Disassembler tools.
• Develop features for binary analysis and instrumentation, including generating call graphs, monitoring program register usage, and patching GPU binaries.
• Collaborate with GPU architecture teams and debugger/profiler developers to integrate new features and enhance product capabilities.
• Coordinate with related teams to ensure software compatibility, robustness, and efficient code generation.
• Engage with customers and partners to gather feedback and contribute to continuous innovation and feature improvements.

Essential Qualifications and Skills

• Bachelor's or Master's degree in Computer Science, Computer Engineering, or related fields.
• At least 8 years of experience in low-level system software development, including a minimum of 6 years specialized in compilers, assemblers, binary utilities, or binary analysis tools.
• Strong analytical abilities coupled with proficient C/C++ programming skills.
• Experience in compiler development areas such as feature implementation, code generation, or compiler infrastructure.
• Solid understanding of Assembly language, processor Instruction Set Architecture (GPU-specific ISA preferred), and processor architectures.
• Familiarity with object file formats like ELF and debug formats such as DWARF.

Preferred Additional Skills

• Knowledge of debugger, profiler tools, binary tools, and linker internals.
• Experience with binary analysis and instrumentation tools such as BOLT.
• Comfort with AI tools used daily for development support, including Cursor, Codex, or Claude.
• Understanding of GPU development environments and compute APIs like CUDA and OpenCL.