Hi everyone,
I've been studying the NVIDIA simulation stack — Isaac Sim, Isaac Lab, Newton, Warp, OSMO, OpenUSD, DSX, and the broader Omniverse ecosystem — and put together two community resources:
Maps the full stack into five layers:
Vertical Workflows DSX · Isaac Sim/Lab · Kit-CAE
Delivery & Orchestration App Streaming · OSMO · Helm
Runtime & App Construction Kit SDK · Omniverse Libraries
Acceleration & Simulation Warp · PhysX · RTX · Newton
Interoperability & Semantics OpenUSD · SimReady
Curated links to official docs, GitHub repos, related awesome lists, and community projects. Isaac Lab/Sim and OSMO feature prominently in both the Robotics vertical and the Orchestration layer.
A documentation-grounded architecture review arguing that NVIDIA's simulation offering is best understood as layered infrastructure. The robotics section covers Isaac Sim → Isaac Lab → Newton, and how OSMO completes the Physical AI pipeline.
Actively seeking co-authors — especially folks with sim-to-real transfer results, Newton vs. MuJoCo benchmarks, or OSMO orchestration experience at scale. See the Call for Co-Authors.
The awesome list is CC0 and PRs are welcome. Happy to take corrections or suggestions!
Hi everyone,
I've been studying the NVIDIA simulation stack — Isaac Sim, Isaac Lab, Newton, Warp, OSMO, OpenUSD, DSX, and the broader Omniverse ecosystem — and put together two community resources:
1. Awesome list: awesome-nvidia-physical-simulation
Maps the full stack into five layers:
Curated links to official docs, GitHub repos, related awesome lists, and community projects. Isaac Lab/Sim and OSMO feature prominently in both the Robotics vertical and the Orchestration layer.
2. Open working paper: From OpenUSD to DSX
A documentation-grounded architecture review arguing that NVIDIA's simulation offering is best understood as layered infrastructure. The robotics section covers Isaac Sim → Isaac Lab → Newton, and how OSMO completes the Physical AI pipeline.
Actively seeking co-authors — especially folks with sim-to-real transfer results, Newton vs. MuJoCo benchmarks, or OSMO orchestration experience at scale. See the Call for Co-Authors.
The awesome list is CC0 and PRs are welcome. Happy to take corrections or suggestions!