Performance

Performance

Web: performance.html

Belt triad (2.0)

./scripts/grok16-toolchain.sh bench-triad

Workload: field-nexus-bench (240 frames, FieldX86 + entropy + wave + NEXUS).

Toolchain	Profile	compile_ms	run wall_ms	binary bytes
host g++	-O3 -march=native	~2575	~3	~27264
g16	belt_1_0	~2377	~3	~22712
g16	belt_2_0	~3708	~5	~22840

• belt_1_0 — matches host runtime; compile ~8% faster
• belt_2_0 — single fabric production belt; 512 slots, chunked redata

Artifact: data/bench/triad-latest.json

Primary metric

export G16_BELT_PROFILE=belt_2_0
./scripts/grok16-toolchain.sh field-bench

Release rebuild

export G16_RELEASE_PROFILE=1
export G16_ENABLE_LTO=1
export G16_BELT_PROFILE=belt_2_0
./scripts/grok16-toolchain.sh rebuild
./scripts/grok16-toolchain.sh profile
G16_ENABLE_PGO=1 ./scripts/grok16-toolchain.sh field-bench

Bench matrix

./scripts/grok16-toolchain.sh bench-all

Profile	Workload
belt_2_0	field-nexus-bench (2.0 default)
belt_1_0	field-nexus-bench baseline
field_opt	field-nexus-bench
ai	ai-matrix-bench
field_compute	field-canvas-kernel
vulkan_rtx	RTX SIMD (gated)

Gates

• test-battery-heavy — release profile bench
• test-battery-belt — 2.0 belt validation