Loading
Life sciences · Preprint
arXiv · September 8, 2026
Early or partial results. Treat as a signal, not a conclusion.
Ostrich is a GPU-accelerated differentiable rigid-body simulator claimed to achieve higher accuracy and faster convergence than existing engines (MJX, Newton Semi-Implicit) at larger timesteps and lower memory cost. The work is presented as a preprint and reports relative performance metrics (speedups, memory ratios) on specific test scenes, but has not undergone peer review and does not address clinical utility or external validation.
Benchmarking and performance evaluation – comparative tool study. Simulated rigid-body robotic systems; no human or animal subjects.. Intervention: Ostrich: a GPU-accelerated differentiable rigid-body simulator using non-smooth Newton iteration for contact and friction, with implicit function theorem differentiation via implicit function theorem. Compared with: MJX (tape-based differentiable engine) and Newton Semi-Implicit (tape-based semi-implicit integrator).
Ostrich resolves hard contacts and friction at timesteps h ~ 0.1 s, holding MuJoCo's sim-to-real accuracy up to 50x larger timestep than baseline engines Warm iteration runs 211x faster than MJX and 4.7x faster than Newton Semi-Implicit on the tested pallet obstacle scene Ostrich differentiates 8,192 parallel worlds on a single 24 GB GPU, sustaining 29x checkpointed MJX's optimization throughput
Safety was not reported in the material analysed. Check the source before drawing any conclusion about harm.
The source did not state who this applies to in practice.
A novel software tool for differentiable simulation with reported performance gains, but presented as a preprint without peer review, independent validation, or clinical/medical application.
As stated by the source record.
Quoted from the source exactly as published.
Graded across the dimensions that decide whether you should act, each from what the source actually supports. There is no single score, and where a dimension was not assessed it says so.
Three properties determine whether a differentiable simulator can drive gradient-based optimization through contact: simulation accuracy, gradient reliability, and per-iteration cost. Tape-based engines such as MJX and Newton Semi-Implicit require timesteps small enough to keep contacts numerically tractable, and their backpropagation memory grows linearly with the number of timesteps T. Surrogate models bound memory by approximating contact away, but the resulting gradients lose the geometry the optimization depends on. We present Ostrich, a GPU-accelerated rigid-body simulator that resolves hard contacts and friction with non-smooth Newton iteration at large timesteps (h ~ 0.1 s), and differentiates the converged residual via the implicit function theorem, reusing the forward Schur complement to compute the adjoint at O(1) memory per timestep. On real-robot trajectories over a pallet obstacle, Ostrich holds MuJoCo's sim-to-real accuracy up to a 50x larger timestep. Its gradients converge from random initializations where MJX descends slowly and Newton Semi-Implicit stalls; a warm iteration runs 211x faster than MJX's and 4.7x faster than Semi-Implicit's. On the same scene Ostrich differentiates 8,192 parallel worlds on a single 24 GB GPU, sustaining 29x checkpointed MJX's optimization throughput; without checkpointing both baselines exhaust memory at far fewer worlds. We close with a gradient-based trajectory optimization demonstration over triangle-mesh terrain across a 10 s horizon, a setting where prior engines either restrict to primitive geometry or face the convergence and memory limits shown above.
Taken from the source record, never inferred. Follow any of these and new work involving them reaches your briefing.