NVIDIA is seeking a Senior Deep Learning Software Engineer for Recipe Pathfinding to revolutionize SW systems for discovering low-precision and sparsity recipes in LLMs. This role focuses on creating infrastructure and tooling to accelerate the discovery of efficiency-gaining transformations in large language models. The position involves working with cutting-edge hardware features on Blackwell, Rubin, and future platforms, spanning all phases of the LLM lifecycle.
The role is heavily focused on coding, infrastructure development, and performance optimization, with direct impact on NVIDIA's internal software systems and products like Megatron-LM and Transformer Engine. The work is critical for minimizing computational costs in production runs that can reach millions of dollars.
As a world leader in accelerated computing, NVIDIA offers an exciting opportunity to work with some of the most forward-thinking professionals in the technology industry. The company has transformed from its origins in computer graphics to become "the AI computing company," with GPUs now powering deep learning algorithms and acting as the brain of computers, robots, and self-driving cars.
The position offers competitive compensation, including a base salary range of $184,000 - $356,500 USD, equity, and comprehensive benefits. The successful candidate will join a team committed to developing next-generation software solutions and will have the opportunity to influence the future of AI computing platforms.