Python 3 versus C++ g++

How the programs are written matters! Always look at the source code.

If the fastest programs are flagged * possible hand-written vector instructions or "unsafe" or naked ffi, does the host language matter? You might be more interested in the less optimised programs — more seconds, less gz source code.

fannkuch-redux
	elapsed	Mean or 95% CI
source	secs	cpu secs	mem	gz	make
* C++ g++ #6	3.33	13.12	5,800	1535	4.64
* C++ g++ #7	14.06	14.04—14.11	2,449	1156	2.95
C++ g++ #5	7.86	30.73—30.82	5,792	987	3.67
C++ g++ #4	8.91	34.99—35.47	5,710	1445	12.14
C++ g++ #2	9.14	35.88—36.26	5,698	1446	11.86
C++ g++ #3	37.89	37.90—37.99	2,351	600	2.70
C++ g++	10.53	41.25—41.55	5,534	1065	3.88
Python 3 #6	943.88	942.63	11,874	391	3.58
Python 3 #4	311.18	1,221.85	28,701	960	5.27
Python 3 #8	2,457.09	2,455.59	11,874	414	5.65
n-body
	elapsed	Mean or 95% CI
source	secs	cpu secs	mem	gz	make
* C++ g++ #0	2.15	2.16	2,417	1933	3.66
* C++ g++ #7	4.09	4.12—4.17	5,124	1815	4.18
C++ g++ #9	5.15	5.17	2,486	1536	3.03
C++ g++ #3	5.33	5.34	2,482	1402	2.85
* C++ g++ #4	5.46	5.46	2,482	1434	2.83
C++ g++ #2	6.02	6.02—6.10	2,490	1524	3.04
C++ g++ #5	6.33	6.35	2,478	1544	2.92
C++ g++	6.62	6.64	2,499	1666	3.15
C++ g++ #6	6.94	6.93—6.98	2,499	1674	2.98
Python 3	372.41	371.23	12,198	1201	5.23
Python 3 #2	442.06	440.87	12,554	1247	5.24
Python 3 #8	529.38	528.16	12,497	1172	4.94
spectral-norm
	elapsed	Mean or 95% CI
source	secs	cpu secs	mem	gz	make
* C++ g++ #6	0.72	2.86	3,981	1050	3.08
* C++ g++ #5	1.33	5.28	5,820	1050	2.86
C++ g++	5.34	5.34	5,292	459	3.75
C++ g++ #7	1.53	6.07	4,071	1284	2.71
Python 3 #6	349.68	348.48	13,660	334	5.44
Python 3 #4	90.37	352.91	34,693	415	5.41
Python 3 #8	386.45	385.07	12,833	334	5.08
Python 3 #7	109.91	421.84	34,578	425	5.30
mandelbrot
	elapsed	Mean or 95% CI
source	secs	cpu secs	mem	gz	make
* C++ g++ #4	0.89	3.46	37,941	3554	5.04
* C++ g++	0.90	3.48	37,626	1797	3.59
* C++ g++ #6	0.98	3.87	37,908	1014	4.36
C++ g++ #0	2.36	9.31	38,281	2752	4.60
C++ g++ #7	4.87	13.30	38,048	749	3.03
C++ g++ #5	13.47	13.49	36,884	596	2.76
C++ g++ #9	3.53	13.59	37,863	732	3.03
C++ g++ #2	16.96	16.96	36,704	700	3.28
C++ g++ #3	27.38	27.38	5,181	421	3.69
Python 3 #7	182.94	723.28	62,173	698	5.52
Python 3 #3	879.41	877.96	45,085	319	5.26
Python 3 #2	237.64	937.87	230,973	790	5.42
Python 3 #5	260.63	1,033.16	61,620	463	5.48
Python 3 #8	1,577.29	1,575.80	45,031	379	4.84
fasta
	elapsed	Mean or 95% CI
source	secs	cpu secs	mem	gz	make
* C++ g++ #9	0.78	1.56	6,214	2758	6.43
C++ g++ #2	1.85	1.85	5,108	1129	2.89
C++ g++	1.93	1.93	5,104	1064	2.98
C++ g++ #3	2.63	2.63	5,284	1311	3.33
C++ g++ #7	0.88	3.06	6,304	2711	4.58
C++ g++ #4	3.34	3.34	2,363	1291	3.03
C++ g++ #5	1.07	3.69—3.73	7,012	2345	3.90
C++ g++ #6	1.08	3.90	7,127	2326	3.72
Python 3 #3	39.03	38.30—47.80	12,218	1007	5.32
Python 3 #2	39.06	38.61—46.82	12,415	1126	5.21
Python 3 #5	27.95	54.20—54.96	1,190,347	1951	5.52
Python 3	57.39	56.94—65.76	12,100	823	5.06
Python 3 #8	153.17	153.78—158.43	11,915	738	5.04
k-nucleotide
	elapsed	Mean or 95% CI
source	secs	cpu secs	mem	gz	make
C++ g++ #2	2.02	6.05—6.54	162,349	1634	7.41
C++ g++ #3	4.19	13.83—14.44	171,622	1280	5.17
C++ g++	7.04	15.17—15.69	778,510	1452	6.55
Python 3 #3	46.55	182.32—187.37	674,185	1970	5.68
Python 3 #4	76.67	233.00—239.32	523,751	663	4.89
Python 3 #8	234.08	235.36—236.89	642,265	525	5.15
Python 3	234.20	235.10—237.16	515,047	499	5.34
Python 3 #9	66.67	236.43—238.45	392,184	614	4.09
reverse-complement
	elapsed	Mean or 95% CI
source	secs	cpu secs	mem	gz	make
* C++ g++ #2	0.76	0.76	514,101	2099	4.30
Python 3 #3	4.96	4.95	523,801	803	5.28
C++ g++	6.55	6.55—6.68	1,006,805	599	3.43
Python 3 #5	4.18	7.46—7.50	2,418,156	960	5.36
Python 3 #2	10.09	10.13—10.24	1,837,285	440	5.32
Python 3	10.43	10.43—10.59	2,040,713	484	5.28
Python 3 #6	7.59	11.07—11.11	2,795,930	821	5.37
Python 3 #4	13.47	13.44—13.88	4,109,324	343	5.05
C++ g++ #5	37.66	37.65—37.69	792,846	704	3.40
C++ g++ #6		Bad Output
C++ g++ #3		Bad Output
binary-trees [ Contentious. Different approaches. ]
	elapsed	Mean or 95% CI
source	secs	cpu secs	mem	gz	make
C++ g++ #7	1.01	3.46—3.62	175,108	896	7.50
C++ g++ #5	1.17	4.12—4.17	208,171	891	3.87
C++ g++ #9	1.76	4.66—4.82	140,571	815	4.05
C++ g++ #6	1.93	4.90—5.09	273,973	851	4.98
C++ g++ #3	2.28	5.06—5.10	273,715	951	4.79
C++ g++ #4	2.22	5.20—5.33	273,977	826	5.16
C++ g++	2.69	9.43—9.52	1,896,727	760	5.34
C++ g++ #2	22.77	22.81—23.05	274,080	514	3.52
Python 3 #2	100.49	100.69—101.09	281,829	344	5.08
Python 3 #4	33.37	121.32—121.68	460,771	481	5.57
Python 3	34.85	123.08—123.98	445,649	597	5.16
Python 3 #5	35.37	125.71—126.12	806,117	669	5.30
Python 3 #3	51.88	148.91—149.87	819,077	1247	5.39
pidigits [ Contentious. Different libraries. ]
	elapsed	Mean or 95% CI
source	secs	cpu secs	mem	gz	make
C++ g++ #4	0.72	0.72	6,496	521	3.45
* C++ g++ #5	0.73	0.73	6,300	804	3.34
* Python 3 #3	1.35	1.28—1.32	20,009	573	5.34
Python 3 #5	1.40	1.36	15,606	369	7.20
Python 3 #4	4.63	4.60	12,718	348	4.88
regex-redux [ Contentious. Different libraries. ]
	elapsed	Mean or 95% CI
source	secs	cpu secs	mem	gz	make
C++ g++ #6	1.14	2.25	285,639	2856	5.06
* Python 3 #2	1.41	2.65	191,300	1411	5.48
Python 3	8.96	16.47—16.75	459,624	515	5.47
C++ g++ #3	12.37	18.03—18.68	222,614	859	12.29
C++ g++		Empty
C++ g++ #2		Empty
C++ g++ #5		Empty
C++ g++ #4		Empty
Python 3	Python 3.13.0
C++ g++	g++ (Ubuntu 14.2.0-4ubuntu2) 14.2.0