AI Infrastructure Software Engineer — CosmosLab

NVIDIA
Shanghai, Beijing, Shenzhen
Workplace: OnsiteFull timeFunction: Software EngineeringEducation: bachelorsSkills: ["Communication","Collaboration","Intellectual curiosity","Problem-solving","Defensive programming"]

Join NVIDIA’s Cosmos Lab Infra team to design and scale AI training infrastructure for physical AI foundation models. You’ll build training, inference, and evaluation pipelines across pre-training, SFT, and RL post-training, including control-plane orchestration across clusters. Work on distributed training backends (sharding, mixed precision, checkpointing), simulation-to-RL training loops, and reliability/efficiency metrics—plus deep debugging across application, framework, GPU, and network layers.

Loading

Loading job details...

Preparing the role view and application actions.

FursaFursa
NVIDIA
NVIDIA
51 minutes ago

AI Infrastructure Software Engineer — CosmosLab

✓ Verified Job

Canonical indexed version, validated from employer's careers page.

Source: Company careers pageValidated by: Fursa AI
Last checked: 51 minutes agoStatus: Live

Job Summary

Join NVIDIA’s Cosmos Lab Infra team to design and scale AI training infrastructure for physical AI foundation models. You’ll build training, inference, and evaluation pipelines across pre-training, SFT, and RL post-training, including control-plane orchestration across clusters. Work on distributed training backends (sharding, mixed precision, checkpointing), simulation-to-RL training loops, and reliability/efficiency metrics—plus deep debugging across application, framework, GPU, and network layers.
Location: Shanghai, Beijing, Shenzhen
Workplace: Onsite
Employment Type: Full time
Job Function: Software Engineering
Seniority: Mid level

Key Responsibilities

  • •Create and implement training infrastructure across pre-training, SFT, and RL post-training, including framework and control-plane coordination across clusters.
  • •Develop and improve pre-training and SFT pipelines for large-scale data loading, distributed training, and checkpointing to reach high throughput and scalability.
  • •Develop and improve inference and evaluation stack, including inference engine, inference/generation pipelines (supporting RL rollout), and evaluation pipelines for low latency/high throughput.
  • •Integrate and orchestrate simulation and robotics environments as RL environments, driving the simulation↔rollout↔training loop at scale.
  • •Build and refine distributed training backend and improve efficiency, scalability, and resiliency, including fault tolerance and failure triage from application through framework/GPU/network/hardware.

Key Requirements

  • •5+ years developing software infrastructure for large-scale AI or distributed systems.
  • •Bachelor’s degree or higher in Computer Science or a related technical field (or equivalent experience).
  • •Strong debugging and triage skills across the stack, from AI application down to GPU/hardware behavior.
  • •Proven track record building and scaling large-scale distributed systems, ideally distributed training or inference.
  • •Hands-on experience with AI training and/or inference infrastructure, including RL/post-training, training frameworks, or inference serving.
Experience:Large-scale AIDistributed systemsAI trainingAI inferenceRL/post-training
Education:Bachelor's in Computer Science or related technical field
Skills:CommunicationCollaborationIntellectual curiosityProblem-solvingDefensive programming
Tech Stack:PythonPyTorchFSDPDTensorMegatronPPOGRPODPOCC++CUDACIDistributed trainingInference serving

Company Brief

NVIDIA
Designs and manufactures GPUs, AI accelerators, and system-on-chip products for gaming, data centers, professional visualization, and automotive markets, enabling advanced graphics, AI, and high-performance computing solutions worldwide.
Industry: Electronics Manufacturing
Company Size: Enterprise (1,001+ employees)
Revenue: USD 1B+
Growth: Public Company
Valuation: Public Company (Market Cap in USD)
Funding: IPO / Publicly Listed
Headquarters: Santa Clara, United States
Founded: 1993
Glassdoor
Glassdoor: 4.3
WebsiteLinkedInGlassdoor