HPC Compiler Developer

NVIDIA•Hillsboro, OR

18h

About The Position

NVIDIA’s HPC compiler group wants to hire a Compiler Developer to join the team building one of the world's most important optimizing compilers. If you want to be in the driver seat of state-of-the-art compiler development for high-performance computing, then we want you! We're the team implementing C++ Standard Language Parallel Algorithms, OpenACC, and Flang with a keen focus on high performance for GPUs and CPUs using MLIR, LLVM, and runtime libraries.

Requirements

Bachelors, Masters, or Ph.D. Degree in Computer Science, ECE or equivalent experience
3+ years experience developing production-quality compilers, libraries, or similar software
Proven understanding of compiler internals including experience targeting auto-parallelization and GPU targets
Knowledge of programming fundamentals, especially in C/C++, with a deep understanding of parallel computer architectures and optimizations

Nice To Haves

Direct experience with Fortran or GPU programming models is a huge plus
Experience writing code using Modern C++
Experience with GPU-based parallel computing
Familiarity with C++, Fortran, OpenACC, OpenMP, or CUDA compilation
You have a real passion for compiler development

Responsibilities

Solve difficult problems at the intersection of compilers, runtime libraries, programming models, and high-performance computing architectures
Implement new and future OpenACC, OpenMP, and standard language parallelism features for C++, Fortran, and Python for NVIDIA GPUs and Multicore CPUs
Participate in the evolution of standard language parallelism for GPUs and Multicore CPUs
Balance individual effort with teamwork throughout debugging, prototyping, and productization
Discuss and refine software designs and implementation strategies with peers
Partner with application teams to investigate interesting and meaningful compilation problems from the HPC community