Python Interpreters Benchmarks
x64 ArchLinux : AMD® Ryzen 7 4700U®

- all benchmarks - n-body fannkuch-redux meteor-contest fasta fasta-redux fibonacci fib50 spectral-norm reverse-complement mandelbrot k-nucleotide regex-dna pidigits chameneos-redux templates thread-ring binary-trees-redux binary-trees pystone richards iobench jsonbench - all languages - Python 2 PyPy Python 3 Python dev PyPy 3 Jython IronPython Cython Nuitka Shedskin Numba Pyston MicroPython Grumpy Graal RustPython

vs Python 2 PyPy Python 3 Python dev PyPy 3 Jython IronPython Cython Nuitka Shedskin Numba Pyston MicroPython Grumpy Graal RustPython x64 octa-core

 1 : Are the PyPy 3 programs faster? At a glance.

Each chart bar shows, for one unidentified benchmark, how much the fastest PyPy 3 program used compared to the fastest Python 3 program.

(Memory use is only compared for tasks that require memory to be allocated.)

These are not the only compilers and interpreters. These are not the only programs that could be written. These are not the only tasks that could be solved. These are just 10 tiny examples.

 2 : Are the PyPy 3 programs faster? Approximately.

Each table row shows, for one named benchmark, how much the fastest PyPy 3 program used compared to the fastest Python 3 program.

(Memory use is only compared for tasks that require memory to be allocated.)

PyPy 3 used what fraction? used how many times more?
Benchmark	Time	Memory	Code
n-body	1/9		±
richards	1/4		±
fib50	1/3		±
meteor-contest	1/3		±
pystone	1/2	?	±
jsonbench	1/2		±
fasta †	1/2		±
fasta-redux	±	?	±
spectral-norm †	±	?	2×
fannkuch-redux	±		±
thread-ring	±		±
k-nucleotide †	2×	?	1/3
iobench	2×		±
binary-trees	3×	2×	±
reverse-complement †	3×	?	±
regex-dna †	4×	?	±
pidigits †	124×	?	±

PyPy 3 used what fraction? used how many times more?
Time-used	|-	|---	25%	median	75%	---|	-|
(Elapsed secs)	1/9	1/9	1/2	±	2×	4×	124×

† possible mismatch - one-core program compared to multi-core program.

± read the measurements and then read the program source code.

 3 : Are the PyPy 3 programs faster? Measurements.

These are not the only tasks that could be solved. These are just 10 tiny examples. These are not the only compilers and interpreters. These are not the only programs that could be written.

For each named benchmark, measurements of the fastest PyPy 3 program are shown for comparison against measurements of the fastest Python 3 program.

Program Source Code	CPU secs	Elapsed secs	Memory KB	Code B	≈ CPU Load
n-body
PyPy 3	3.19	3.19	67,464	1315	0% 0% 100% 0% 0% 1% 0% 1%
Python 3	28.83	28.83	8,844	1315	0% 0% 0% 0% 0% 0% 100% 0%
richards
PyPy 3	0.18	0.18	?	2434	0% 0% 0% 100% 0% 6% 0% 10%
Python 3	0.70	0.70	9,424	2434	0% 0% 0% 0% 0% 100% 0% 1%
fib50
PyPy 3	0.59	0.59	70,684	136	2% 0% 100% 0% 0% 2% 0% 2%
Python 3	2.04	2.04	8,348	136	2% 0% 0% 0% 0% 0% 100% 0%
meteor-contest
PyPy 3	0.64	0.65	74,028	1443	3% 2% 100% 2% 0% 6% 2% 3%
Python 3	2.09	2.10	9,528	1443	1% 0% 0% 0% 0% 0% 0% 100%
pystone
PyPy 3	0.08	0.09	?	2300	0% 0% 88% 0% 0% 12% 11% 0%
Python 3	0.19	0.19	?	2300	5% 5% 0% 0% 0% 100% 0% 0%
jsonbench
PyPy 3	0.96	0.97	76,696	322	0% 0% 1% 0% 0% 0% 100% 2%
Python 3	2.14	2.15	13,136	322	0% 2% 0% 1% 0% 1% 100% 1%
fasta
PyPy 3	1.36	1.37	70,148	1698	4% 1% 2% 1% 3% 100% 2% 2% †
Python 3	4.80	2.66	63,220	2016	39% 6% 54% 3% 7% 41% 32% 5% †
fasta-redux
PyPy 3	0.15	0.15	?	1115	6% 100% 0% 0% 0% 12% 0% 0%
Python 3	0.19	0.19	?	1115	100% 0% 0% 0% 0% 5% 0% 0%
spectral-norm
PyPy 3	0.09	0.09	?	594	0% 18% 100% 0% 11% 0% 0% 11% †
Python 3	0.10	0.10	?	394	0% 92% 0% 0% 0% 0% 0% 0% †
fannkuch-redux
PyPy 3	1.72	0.67	80,080	1008	14% 15% 55% 53% 47% 40% 46% 1%
Python 3	5.43	0.73	89,420	894	88% 92% 93% 93% 95% 97% 93% 93%
thread-ring
PyPy 3	28.61	37.17	73,036	448	19% 7% 4% 3% 12% 24% 5% 15%
Python 3	20.45	38.97	11,040	448	4% 8% 5% 5% 4% 7% 3% 11%
k-nucleotide
PyPy 3	0.10	0.10	?	594	0% 100% 0% 0% 0% 9% 0% 0% †
Python 3	0.12	0.06	?	2011	17% 20% 20% 83% 33% 17% 20% 20% †
iobench
PyPy 3	5.29	5.29	74,204	367	100% 0% 0% 1% 0% 0% 2% 0%
Python 3	2.84	2.84	9,080	367	2% 2% 0% 100% 0% 1% 1% 1%
binary-trees
PyPy 3	2.28	0.90	128,080	741	63% 24% 24% 26% 27% 27% 36% 37%
Python 3	1.52	0.29	80,548	741	67% 69% 66% 64% 63% 59% 67% 79%
reverse-complement
PyPy 3	0.06	0.06	?	458	0% 0% 0% 0% 0% 0% 0% 100% †
Python 3	0.02	0.02	?	449	0% 33% 100% 0% 0% 0% 0% 0% †
regex-dna
PyPy 3	0.16	0.16	?	524	100% 6% 0% 0% 12% 0% 6% 0% †
Python 3	0.04	0.04	?	524	0% 0% 100% 0% 0% 0% 0% 0% †
pidigits
PyPy 3	2.84	2.84	75,604	322	76% 2% 0% 0% 1% 24% 1% 0% †
Python 3	0.02	0.02	?	379	25% 0% 0% 33% 0% 0% 0% 100% †
mandelbrot
	No program
Python 3	37.12	4.84	294,184	3852	96% 98% 96% 94% 94% 98% 97% 97%
templates
	No program
Python 3	0.63	0.64	18,364	322	4% 0% 0% 100% 0% 0% 0% 0%
fibonacci
PyPy 3	0.16	0.16	?	182	94% 0% 6% 6% 0% 0% 0% 0%
	No program
chameneos-redux
	No program
Python 3	182.63	264.55	14,784	1191	5% 3% 3% 2% 12% 11% 12% 11%
binary-trees-redux
	No programs

† possible mismatch - one-core program compared to multi-core program.

 4 : Are there other PyPy 3 programs for these benchmarks?

Remember - those are just the fastest PyPy 3 and Python 3 programs measured on this OS/machine. Check if there are other implementations of these benchmark programs for PyPy 3.

Python 2 PyPy Python 3 Python dev PyPy 3 Jython IronPython Cython Nuitka Shedskin Numba Pyston MicroPython Grumpy Graal RustPython

Maybe one of those other PyPy 3 programs is fastest on a different OS/machine.

 5 : Are there other faster programs for these benchmarks?

Remember - those are just the fastest PyPy 3 and Python 3 programs measured on this OS/machine. Check if there are faster implementations of these benchmark programs for other programming languages.

n-body fannkuch-redux meteor-contest fasta fasta-redux fibonacci fib50 spectral-norm reverse-complement mandelbrot k-nucleotide regex-dna pidigits chameneos-redux templates thread-ring binary-trees-redux binary-trees pystone richards iobench jsonbench

Maybe one of those other programs is fastest on a different OS/machine.

 PyPy 3 : the new Python JIT 

  Home   Conclusions   License   Play