Serving World Models: Latency vs Throughput with NVIDIA Cosmos 3 Super on Nebius | Cosmos Labs @NVIDIADeveloper
Serving World Models: Latency vs Throughput with NVIDIA Cosmos 3 Super on Nebius | Cosmos Labs  @NVIDIADeveloper
Uploaded October 2026 | Updated November 2026, 4 days ago
Serving world models for video generation requires a different balance of latency and throughput than serving large language models. In this livestream, Nebius Physical AI team members Stewart Tong and Timothy Le examine NVIDIA Cosmos 3 Super benchmarks through vLLM-Omni on NVIDIA HGX B200 and HGX H200 systems.

Their work compares four serving topologies, ranging from one eight-GPU replica to eight single-GPU replicas. See how replica count, parallelism, and concurrency affect request latency and validated video-seconds per node-hour. Learn which settings to hold constant for video generation benchmarks and how to check generated clips for file integrity and basic motion.

What You'll Learn:
- Why faster video requests can mean less output per node
- How to measure latency and throughput with video-validation checks
- When more replicas help and extra concurrency adds delay
- What NVIDIA HGX B200 and HGX H200 benchmarks show about allocating GPUs
- Have questions about serving NVIDIA Cosmos 3 Super for world-model video generation? Drop them in live β€” the NVIDIA and Nebius teams will address them during the discussion.

Learn more about NVIDIA Cosmos 3 serving and reproduce the benchmark:

πŸ“„ Read the technical blog β†’ nebius.com/blog/posts/serving-world-models-latency-node-throughput
πŸ§ͺ Reproduce the benchmarks β†’ github.com/nebius/nebius-physical-ai
πŸ“š Explore the NVIDIA Cosmos platform β†’ github.com/nvidia/Cosmos
πŸ‘₯ Join the NVIDIA Omniverse community β†’ discord.com/invite/nvidiaomniverse
Serving World Models: Latency vs Throughput with NVIDIA Cosmos 3 Super on Nebius | Cosmos LabsA chess grandmaster entered our hackathon. Heres what he built.Ask the Experts: How NVIDIA OpenShell Secures Autonomous Agents | Nemotron LabsSimplicits: Mesh-Free, Geometry-Agnostic, Elastic Simulation | NVIDIA ResearchBreaking Down The Agent StackCosmos | SIGGRAPH 2026 RecapRun Your Polars Code on Multiple GPUs | Live with cuDF PolarsPost-Train NVIDIA Cosmos 3 In a Day with NVIDIA TAO Agent Skills | Cosmos LabsTurn Research Papers into Insights with DeepSeek-V4 and SGLangBeyond VLAs: How World Action Models Reshape Robot ManipulationJETSON AI LAB | Research Group Meeting (6/25/2024)How to Maximize Memory Efficiency and Reduce Costs on NVIDIA Jetson
NVIDIA Developer |

Serving World Models: Latency vs Throughput with NVIDIA Cosmos 3 Super on Nebius | Cosmos Labs

SHARE TO X SHARE TO REDDIT SHARE TO FACEBOOK WALLPAPER