We are now looking for a Senior Software Engineer for Quantized Training. We are a team committed to developing next-generation quantized training recipes for Hopper and future GPUs. We are seeking software engineers to help rethink and create tailored solutions to accelerate the discovery of new recipes. This is a coding-heavy role focused on building infrastructure, tooling, and visualizations.
The candidate's work directly supports NVIDIA's production SW systems including Megatron-LM and Transformer Engine. The candidate will be part of a core team of engineers and researchers working in lock step to improve quantized training convergence and efficiency.
What you'll be doing:
What we need to see:
Ways to stand out from the crowd:
GPU computing is the most productive and pervasive platform for deep learning and AI. NVIDIA offers highly competitive salaries and a comprehensive benefits package. This opportunity offers you the ability to collaborate with some of the most forward-thinking and hard-working people in the world, shaping the future of AI in a creative and autonomous work environment that encourages innovation.