Senior Software Engineer, Cosmos Infrastructure and End to End Performance

NVIDIA Gruppe; Santa Clara (CA); On-site; USD 184,000 - 288,000; Full time

Benefits offered by this job

Equity

Benefits

Job summary

NVIDIA Cosmos is recruiting a Senior Software Engineer to advance infrastructure for world foundation models and end-to-end performance. You will optimize large-scale AI workloads, shape distributed systems for data centers and edge deployments, and collaborate with customers to simplify usage and expand the ecosystem.

Bring a Master’s in a related field and 5+ years of experience, with strong Python/C/C++, Distributed PyTorch, and CUDA skills. Expect an equity package and comprehensive benefits.

Qualifications

• Masters in Computer Engineering, CS, EE or related STEM with 5+ years of experience.
• Expertise with large scale parallel and distributed accelerator-based systems.
• Expertise optimizing AI workloads and performance modeling at scale.
• Proficiency in Distributed PyTorch; Python, C/C++. Strong background in Computer Architecture, Networking, Storage, Accelerators.
• Understanding of DNNs and their use in AI/ML applications.
• Experience with at least one public CSP (GCP AWS Azure OCI).
• Experience automating multimodal data ingestion and curation; tokenization and dataset prep is a plus.
• Experience building transformer models or RL fine-tuning and post-training frameworks.
• Understanding optimization for inference - export, quantization and containerization.

Responsibilities

• Build SOTA world foundation models (e.g., Cosmos3).
• Drive end-to-end performance analysis and HW-SW codesign for data center and edge deployments.
• Engage with customers to ensure Cosmos models are easy to use and ecosystem-friendly.
• Develop infrastructure to automate data ingestion, curation, pre-training, post-training, export/quantization, and edge deployment.
• Design for robustness and fault tolerance.

Skills

Python

C/C++

Distributed PyTorch

CUDA

Computer Architecture

Networking

Storage systems

Performance optimization

DNNs

Public CSP infrastructure (GCP AWS OCI

Education

Master's degree in Computer Engineering / Computer Science / Electrical Engineering

Tools

PyTorch

TensorFlow

Megatron-LM

JAX

Cosmos

Tensort-LLM

VLLM

Job description

We're now looking for a Senior Software Engineer, Cosmos Infrastructure and End to End Performance! NVIDIA Cosmos is an open omni-model platform of generative world foundation models (WFMs) designed to accelerate physical AI. By combining world generation, physical reasoning, and action generation into unified systems, Cosmos helps developers simulate physical environments and train robots, autonomous vehicles, and smart spaces. You can explore the project via the NVIDIA Cosmos GitHub or look up releases on Hugging Face.

What you will be doing:

• Building SoTA, World foundation models (https://arxiv.org/pdf/2606.02800) (like Cosmos3);
• Engage with driving end to end performance analysis and drive HW-SW codesign for both the data center infrastructure and Edge deployments
• Engage with customers to ensure the Cosmos models are easy to use and enabling the ecosystem;
• Develop infrastructure to improve and automate the entire process of data ingestion, curation, pre-training, post-training, Export/quantization and deployment on the edge;
• Design for robustness and fault tolerance.

What we need to see:

• A Masters in Computer Engineering, Computer Science, Electrical Engineering or related STEM degree or equivalent experience. 5 years of relevant work experience
• Expertise in working with large scale parallel and distributed accelerator-based system systems
• Expertise optimizing performance and AI workloads on large scale systems. Experience with performance modeling and benchmarking at scale
• Proficiency in Distributed PyTorch; Python, C/C++. A strong background in Computer Architecture, Networking, Storage systems, Accelerators
• Understanding of DNNs and their use in emerging AI/ML applications and services
• Expertise with at least one of public CSP infrastructure (GCP, AWS, Azure, OCI, …)
• A deep understanding of World Foundation Models and their application to Physical AI
• Experience developing infrastructure to automate multimodal data ingestion and curation. Experience driving tokenization and data set preparation is a plus.
• Prior experience building transformer models (autoregressive and diffusion) or building the framework for and driving Post training - fine tuning and RL algorithms
• Understanding of how to optimize for inference - export, quantization and containerization

Ways to stand out from the crowd:

• Familiarity with popular AI frameworks (TensorFlow, JAX, Cosmos, Megatron-LM, Tensort-LLM, VLLM) among others.
• Proficiency in CUDA
• Very high intellectual curiosity; Confidence to dig in as needed; Not afraid of confronting complexity; Able to pick up new areas quickly with excellent interpersonal skills

Your base salary will be determined based on your location, experience, and the pay of employees in similar positions. The base salary range is 152,000 USD - 241,500 USD for Level 3, and 184,000 USD - 287,500 USD for Level 4.

You will also be eligible for equity and benefits.

Applications for this job will be accepted at least until September 12, 2026.

This posting is for an existing vacancy.

NVIDIA uses AI tools in its recruiting processes.

NVIDIA is committed to fostering an inclusive work environment and proud to be an equal opportunity employer. As we highly value diversity in our current and future employees, we do not discriminate (including in our hiring and promotion practices) on the basis of race, religion, color, national origin, gender, gender expression, sexual orientation, age, marital status, veteran status, disability status or any other characteristic protected by law.