Senior Software Engineer, Cosmos Infrastructure and End to End Performance

We are seeking a Senior Software Engineer to join our NVIDIA Cosmos team, which focuses on building an open omni-model platform of generative world foundation models (WFMs) designed to accelerate physical AI. The successful candidate will work on building and optimizing large-scale parallel and distributed systems, driving end-to-end performance analysis, and collaborating with customers to ensure ease of use.

Key Responsibilities:

• Building State-of-the-Art (SoTA) world foundation models, such as Cosmos3
• Engaging in end-to-end performance analysis and HW-SW co-design for data center infrastructure and Edge deployments
• Collaborating with customers to ensure ease of use and enablement of the ecosystem
• Developing infrastructure to improve and automate data ingestion, curation, pre-training, post-training, export/quantization, and deployment on the edge
• Designing for robustness and fault tolerance

Requirements:

• Master's degree in Computer Engineering, Computer Science, Electrical Engineering, or related STEM field, or equivalent experience
• 5+ years of relevant work experience
• Expertise in large-scale parallel and distributed accelerator-based systems
• Proficiency in Distributed PyTorch, Python, C/C++, and computer architecture, networking, storage systems, and accelerators
• Understanding of DNNs and their applications in emerging AI/ML services
• Experience with public CSP infrastructure (GCP, AWS, Azure, OCI, etc.)
• Familiarity with world foundation models and their applications to Physical AI

Nice to Have:

• Familiarity with popular AI frameworks (TensorFlow, JAX, Cosmos, Megatron-LM, etc.)
• Proficiency in CUDA
• High intellectual curiosity and excellent interpersonal skills

Compensation:

• Base salary range: $152,000 - $241,500 (Level 3) or $184,000 - $287,500 (Level 4)
• Equity and benefits package