rayrai_benchmark

rayrai_benchmark measures rayrai rendering and serialization paths, including shadowed and unshadowed color rendering, readback, depth, picking, transparent visuals, culling, instanced visuals, point clouds, quality presets, HDR/IBL setup, scene synchronization, and TCP deformable serialization.

Run:

<raisim-install>/bin/rayrai_benchmark
<raisim-install>/bin/rayrai_benchmark --json
<raisim-install>/bin/rayrai_benchmark --only=viewer_quality_fast --frames=60 --warmup=10
<raisim-install>/bin/rayrai_benchmark --objects=1024 --width=1280 --height=720

Supported options are --objects=, --frames=, --warmup=, --width=, --height=, --min-lod-speedup=, --only=, and --json.

When Python support is available, the rayrai_benchmark_json custom target can run the benchmark, check thresholds, and write the configured JSON path.

For focused Rayrai timing, use rayrai_complete_showcase_benchmark for the full sensor showcase and rayrai_pbr_first_draw_benchmark for PBR shader-material first-draw cost.