Interpreter benchmarks

All workloads

Workload	piko median	Go	CPython	PyPy	tengo	scriggo	mvm	yaegi
Smoke: integer addition	444 µs	0.06×	3.44×	1.08×	11.4×	10.1×	6.33×	5.09×
Iterative Fibonacci	6.58 ms	0.02×	2.60×	1.23×	unsupported	2.32×	5.13×	4.27×
Word frequency	182 ms	0.10×	1.72×	0.22×	4.32×	2.29×	3.82×	2.54×
Levenshtein, 1k pairs	191 ms	0.03×	1.03×	0.13×	2.84×	1.89×	3.95×	3.45×
Mini JSON parse	208 ms	0.06×	0.89×	0.48×	unsupported	unsupported	2.09×	10.3×
Expression evaluator	2.15 s	0.04×	0.66×	0.10×	2.02×	unsupported	2.42×	5.93×
LRU cache	155 ms	0.10×	0.64×	0.10×	1.13×	unsupported	2.09×	6.25×
Dijkstra shortest paths	112 ms	0.04×	0.80×	0.33×	1.43×	1.31×	unsupported	2.60×
Sudoku solver, 100 boards	20.9 ms	0.05×	1.27×	1.02×	2.89×	unsupported	2.42×	6.60×
Game of Life	4.64 s	0.02×	1.29×	0.03×	2.31×	1.44×	unsupported	2.26×
Markov text generation	23.7 ms	0.23×	1.14×	0.40×	2.65×	2.37×	2.64×	2.46×
Trie of 50k words	527 ms	0.02×	1.04×	0.10×	2.79×	unsupported	2.43×	2.51×
Brainfuck interpreter	294 µs	0.10×	1.32×	10.4×	60.3×	1.72×	8.15×	5.85×
Polymorphic AST eval	54.9 ms	0.04×	1.01×	1.46×	2.18×	unsupported	4.49×	25.5×
Mandelbrot (FP)	28.8 ms	0.12×	4.86×	0.38×	9.36×	6.93×	10.1×	8.24×
Open-addressing hash map	506 ms	0.04×	1.33×	0.18×	3.39×	unsupported	2.08×	3.49×
Parallel word count (Monte Cristo)	108 ms	0.18×	1.68×	3.28×	unsupported	unsupported	2.80×	unsupported
Closures pipeline	150 ms	0.02×	0.77×	0.15×	1.62×	0.99×	1.20×	5.40×
Generic numeric pipeline	26.9 ms	0.03×	1.26×	0.36×	3.38×	unsupported	2.07×	1.57×
Type switches	75.1 ms	0.01×	1.73×	0.36×	2.04×	1.70×	1.94×	4.22×
Invert binary tree	116 ms	0.05×	0.79×	0.34×	1.62×	unsupported	1.57×	6.41×
Dense neural-network layer	1.10 ms	0.13×	20.5×	5.19×	31.2×	12.0×	22.0×	15.3×
N-body solar system	533 ms	0.02×	0.74×	0.05×	2.14×	1.80×	2.17×	2.92×
Per-line SHA-256 (FFI overhead)	39.3 ms	0.05×	1.46×	0.80×	2.23×	2.10×	3.16×	1.42×

Interpreter benchmarks live data