Performance

Performance

Web: performance.html · speed-bench.html

Speed bench (3.0 — report v3.0.0)

Suite: speed_demo @ 1.0.0 · 3s execution window · Schema: grok16-field-exec-full-bench/v3

SPEED_DEMO_TARGET_SEC=3 ./scripts/grok16-toolchain.sh exec-full-bench
./scripts/grok16-toolchain.sh exec-compare

Runner	Compile (ms)	ops/s
C++ g16 belt_2_0	2,494	85.3M
C++ host g++ -O2	1,710	83.2M
CMake host g++ -O2	3,682	82.6M
C g16 belt_2_0	318	79.5M
C host gcc -O2	347	73.4M
Python host CPython 3	—	778K
Python gpy-16 GrokVM	—	766K

Artifacts: docs/SPEED-BENCH-REPORT.md · docs/field-exec-full-bench.json · Wiki: Speed-Bench

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

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

Gates

• test-battery-heavy — release profile bench
• test-battery-belt — 2.0 belt validation
• exec-full-bench — 3.0 compile + execution speed gate