Parallel word count (Monte Cristo)

Parallel word-frequency count over ~2.7 MB of Project Gutenberg text. Each runner uses its idiomatic CPU-parallel primitive.

Compile time · median (cold)

median of 7 runs

Native Gocompiled

113 ms34.3×

Pipitbytecode VM

3.29 msbaseline

CPython 3.14bytecode VM

843 µs0.26×

PyPy 7.3tracing JIT

602 µs0.18×

Ttengobytecode VM

no data

n/a

Sscriggobytecode VM

unsupported in scriggo

n/a

Mmvmbytecode VM

800 µs0.24×

YyaegiAST walker

no data

n/a

Runner	N	Compile	Runtime	P95	Stddev	RSS	vs pipit	Status
Native Gocompiled	7	113 ms	22.9 ms	24.2 ms	800 µs	1.62 GiB	34.3×	OK
Pipitbytecode VM	7	3.29 ms	90.4 ms	93.7 ms	1.50 ms	592 MiB	1.00×	OK
CPython 3.14bytecode VM	7	843 µs	190 ms	193 ms	1.45 ms	n/a	0.26×	OK
PyPy 7.3tracing JIT	7	602 µs	298 ms	299 ms	1.78 ms	n/a	0.18×	OK
tengobytecode VM	0	n/a	n/a	n/a	n/a	n/a	n/a	no data
scriggobytecode VM	0	n/a	n/a	n/a	n/a	n/a	n/a	unsupported
mvmbytecode VM	7	800 µs	256 ms	266 ms	9.14 ms	1.69 GiB	0.24×	OK
yaegiAST walker	0	n/a	n/a	n/a	n/a	n/a	n/a	no data

Workload & symmetry rules

Workload

Load The Count of Monte Cristo (Project Gutenberg #1184) corpus from disk.
Split into 16 word-aligned chunks.
Spawn 16 workers (goroutine + WaitGroup on Go/pipit; multiprocessing.Pool on Python).
Each worker tokenises via byte-walk and builds a local word → count map.
Merge maps, compute top-50 by (count desc, word asc) via hand-rolled insertion-sort.

Each runner uses its idiomatic CPU-parallel primitive. Python's multiprocessing.Pool (not threading, because of the GIL).
No str.split, no regexp, no sort.*/sorted(key=...): tokenisation and ranking are hand-rolled.

Directly measures the GIL-free advantage. Pipit inherits Go's goroutine scheduler; CPython has to fork.

Source code