Nvidia Senior HPC Performance Engineer - AI Science Scale 

United States, Texas 

206678507

What you'll be doing:

• Design and implement computationally performant features for large scale, CUDA-backed ML training frameworks, using low level acceleration and scaling strategies such as GPU porting, data structure innovations, distributed learning technologies

• Optimize computational performance of wide range of business-critical ML models via accelerated hardware and software stack, as well as algorithmic improvements

• Develop and maintain HPC software stack for generative machine learning models in digital biology and beyond

• Collaborate with multiple HPC, AI infrastructure, and research teams

• Develop tools to assist data processing, data quality control, algorithm development, and algorithm testing

• Drive the testing and maintenance of the algorithms and software modules
  

What we need to see:

• Advanced degree in a quantitative field such as Computer Science, Computational Biophysics, Computational Chemistry, Physics, Mathematics, or equivalent experience

• 8+ years of relevant experience

• Consistent track record in performance engineering as well as software design, building and packaging and launching software products

• Deep understanding of parallel programming in C++, Python; ideally CUDA programming experience

• Fluent in modern machine learning frameworks such as PyTorch, TensorFlow, JAX, Warp

• Experience with HPC solutions to research problems, ideally for biology, chemistry or material science applications

• Recognized for technical leadership contributions, capable of self-direction, and ability to learn from and teach others

• You should display strong communication skills, be organized and self-motivated, and play well with others (be an excellent teammate!)
  

Ways to stand out from the crowd:

• Contributor to major scientific AI for Science codebase

• Familiarity with pioneering language and geometric models used in AI for Science applications in biology, chemistry, material science