Senior Deep Learning Software Engineer, Recipe Pathfinding

Details

We are now seeking a Senior Deep Learning Software Engineer for Recipe Pathfinding. NVIDIA is looking for experienced software engineers to help rethink and create software systems to accelerate the discovery of new low-precision and sparsity recipes. A recipe defines which operators in a large language model (LLM) are transformed into low-precision and/or sparsified variants, thereby unlocking efficiency gains.

Recipes can be statically defined at model load time or dynamically adapt to a layer’s input distribution. Recipes can incorporate and compose algorithmic techniques like rotations or low-rank decompositions to tame issues such as outliers. The team develops next-generation software to utilize novel hardware features on Blackwell, Rubin, and beyond. The scope spans all LLM lifecycle phases: pre-training, post-training, and generation. This is a coding-heavy role focused on infrastructure, tooling, and performance, with a goal to minimize production run costs.

Responsibilities

• Create well-designed and well-tested software systems and prototypes to support recipe exploration for research.
• Analyze and prototype state-of-the-art methods for quantization and sparsity.
• Benchmark, profile, and optimize LLM workloads in cluster settings.
• Improve developer and researcher productivity by removing obstacles such as slow CI and training systems.
• Participate in code reviews and address feedback.

Requirements

• PhD or MS degree, or equivalent experience in Computer Science or related field.
• 5+ years of experience in relevant software engineering roles.
• Strong software engineering background with focus on concise and well-tested code.
• Proficient in Python.
• Experience working with ML accelerators.
• Experience with PyTorch or similar frameworks.
• Skills in performance optimization and debugging.
• Strong written and oral communication skills.

Preferred Qualifications

• Proficiency in precision and numerics for machine learning.
• Experience with CUDA and C++ programming.
• Solid foundation in LLM pre-training, post-training, or generation.

About NVIDIA

NVIDIA GPUs power deep learning algorithms, acting as the brain of AI systems including computers, robots, and self-driving cars. NVIDIA is known as the AI computing company and is considered one of the technology world’s most desirable employers, fostering a diverse and inclusive environment.

Benefits

You will be eligible for equity and benefits. The company is committed to equal opportunity and diversity in the workplace.