NVIDIA Warp and MJWarp Enable 2,048 Parallel Robotics Simulations
- Published
- Sep 23, 2026 — 18:41 UTC
NVIDIA's Warp framework now supports a maximum of 2,048 parallel MJWarp environments for robotics simulations. MJWarp implements MuJoCo's physics pipeline on NVIDIA Warp, enhancing simulation capabilities. The SO-101 follower arm model, used in these simulations, requires specific parameters for task success: a maximum horizontal center error of 0.015 m, a minimum vertical separation of 0.035 m, and a maximum vertical separation of 0.055 m. The simulations operate at a controller rate of 50 Hz, with 10 physics steps per control frame and a physics timestep of 0.002 seconds. The SO-101 profile starts with a maximum contacts capacity (nconmax) of 128 and maximum constraints capacity (njmax) of 300. GPU acceleration allows for batch advancements in simulation and learning data, keeping them close to the device. This follows the release of Warp 1.15, which introduced deterministic execution, further improving simulation reliability.
By Callan Zhang · Sep 23, 2026 · Editorial standards →
Summarised from the primary source with AI assistance under human editorial oversight. Turing Wire is not a primary source — read the original for the authoritative account.
Source: Hugging Face Blog
