Software Intern, Compiler

This Software Intern position at NVIDIA involves developing and optimizing kernel-level code using RISC-V Vector Extension (RVV) for NVIDIA GPU and Tegra SOC. The intern will implement vectorized kernels for computational patterns, collaborate with hardware architects and compiler teams, and profile performance. Strong C/C++ systems programming skills, understanding of computer architecture, memory management, and Linux development are essential.

Must Have

• Develop and optimize kernel-level code using RISC-V Vector Extension (RVV) instructions
• Implement vectorized kernels for common computational patterns
• Collaborate with hardware architects and compiler teams
• Write and maintain unit tests and benchmarks
• Profile and analyze kernel performance
• Strong C/C++ systems programming skills
• Understanding of computer architecture and memory management
• Experience with Linux development and build systems (CMake)

Good to Have

• Experience with LLVM compiler

We are a team responsible for building embedded RISC-V CPU for NVIDIA GPU and Tegra SOC, we are looking for software engineering intern for RISC-V Vector (RVV) kernel development

What you'll be doing:

• Develop and optimize kernel-level code using RISC-V Vector Extension (RVV) instructions to accelerate compute-intensive operations
• Implement vectorized kernels for common computational patterns including mathematical operations, data transformations, and memory operations
• Collaborate with hardware architects and compiler teams to understand RVV microarchitecture and optimize for performance
• Write and maintain unit tests and benchmarks to validate correctness and measure performance improvements
• Profile and analyze kernel performance using hardware counters and simulation tools

What we need to see:

• Strong C/C++ systems programming skills
• Understanding of computer architecture and memory management
• Experience with Linux development and build systems (CMake)
• Experience with LLVM compiler is preferred

What You'll Learn:

• RISC-V Vector ecosystem and hardware-software interfaces
• High-performance computing optimization techniques
• Vector processing architectures