New MicroPython Pi calculation performance comparison

dcexpert

New MicroPython Pi calculation performance comparison [Copy link]

 

Last weekend, I did a new comparison of pi calculations, comparing the performance of different processors by calculating pi with different digits. In addition to MicroPython, I also did a comparison of circuitPython calculations.

chip (borad)	freq	ram	ver	100	500	1000	2000	5000	10000	100000
STM32F073 (Nucleo F073RZ)	32	14400	1.18	101	1516	5994	24577	-	-	-
STM32F091 (Nucleo F091RC)	48	24320	1.18	70	1017	4013	15308	-	-	-
STM32F411 (pybnano v2)	96	99328	1.18	16	171	593	2574	16494	58518	-
STM32F411 (pybnano v2) circuitpython	96	86272	7.2.0	16	153	507	1836	16711	71279	-
STM32F405 (pybv10)	168	99456	1.18	10	106	378	1674	10481	36866	-
STM32F413 (Nucleo F413ZH)	96	230784	1.18	14	160	581	2308	17872	65296	-
STM32F429 (Discovery F429)	168	8196480	1.18	15	175	619	2306	14501	59068	13200142
STM32F746 (Nucleo F746ZG)	216	165120	1.18	6	54	179	683	4577	16512	-
STM32F746 (Discovery F7)	192	8196480	1.18	7	60	194	694	4344	17971	3581969
STM32F767 (Nucleo F767ZI)	216	277056	1.18	5	48	163	622	4673	17073	-
STM32F769 (Discovery F769)	216	273024	1.18	5	50	167	634	4776	17729	-
STM32H743 (Nucleo H743ZI)	400	460948	1.18	2	27	92	342	2435	105098	947708
STM32L432 (Nucleo L432KC)	32	50048	1.18	51	547	2019	8494	45673	175622	-
STM32L452 (Nucleo L452RE)	80	139200	1.18	20	209	738	3093	21270	73707	-
STM32L476 (pybL)	80	87360	1.18	twenty one	215	751	3487	20213	71496	-
RP2040 (rpi pico)	125	192064	1.18	16	189	667	2727	21231	74879	-
RP2040 (rpi pico) circuitpython	125	209472	7.2.0	15	158	542	2015	13260	71893	-
ESP8266	80	37952	1.18	44	490	1762	6432	37556	-	-
ESP32	160	111168	1.18	16	124	391	1654	10637	35934	-
ESP32 - PSRAM	160	4098240	1.16	16	192	659	2427	15108	68429	14658189
ESP32-S2	160	128000	1.18	14	124	408	1699	12257	40578	-
ESP32-S2 (Ai-Thinker ESP-12K) circuitpython	240	8195840	7.2.0	9	116	413	1565	9702	43930	22026100
ESP32-S3
ESP32-C3	160	128000	1.18	7	77	280	1332	8838	31028	-
nRF51822 (microbit v1)	16	10048	1.9.2	184	2654	10393	-	-	-	-
nRF52833 (microbit v2)	64	64512	1.15	27	322	1160	4971	30269	113355	-
nRF52840 (PC10056)	64	246528	1.18	29	332	1210	5397	44434	168053	-
nRF52840 (Feather nRF52840 Express) circuitpython	64	142784	7.2.0	twenty three	229	741	2645	18209	84670	-
SAMD21 (circuitplay ground) circuitpython	48	17984	7.2.0	71	906	3463	16974	-	-	-
SAMD51 (PyGamer) circuitpython	120	154112	7.2.0	75	203	537	1787	11396	52945	-
iMXRT1010 (MIMXRT1010 Eval Kit)	500	64000	1.18	3	26	88	363	2062	7927	-
iMXRT1010 (MIMXRT1010 Eval Kit) circuitpython	500	41536	7.2.1	7	64	218	788	5723	23972	-
iMXRT1062 (Teensy 4.0)	600	768348	1.18	2	18	63	236	1668	7923	706754
iMXRT1062 (Teensy 4.0) circuitpython	600	937855	7.2.0	3	40	113	400	2500	15823	1913130
W600	80	54912	1.18	18	228	942	3896	21174	82824	-

illustrate

• Bold represents CircuitPython, normal represents MicroPython
• When repeating calculations, the calculation time will fluctuate slightly, which is normal.
• In order to reduce the interference of other codes on the calculation, clear the original program and reset the hardware before calculation.
• The frequency unit is MHz
• The time unit is ms

For complete test results, please refer to the repository on github

• https://github.com/micropython-Chinese-Community/PiCalcTest