Raspberry Pi Pico VS ESP32 S2 speed comparison benchmark using CircuitPython

Very informative video 😊

How did the pico perform at max overclock (280 MHz)? Default is 125.

Damn nice job, thanks!

The question is how efficient and durable the two devices and compared, ESP32 is erratic, especially triggering its interrupt. I think most of us notice that :)

the rp2040 is around 0.70$ - 0.80$ and can be overclocked it has less IO if i remember correctly but has 2 cores, is pretty fast and cheap

well not really
pico have two core so we can do
133hz*2=266 and is bigger than the esp

so on multi tasking pico is faster in theory

python is monocore so probably not the best choice (maybe cirquitpython is mulricore ?)

there is a Pico version with WiFi. WHy did you say Pico has none?

Это не сравнение. Это какая-то ерунда.

Very nice comparisons you made, comparing boards and libraries (CircuitPy vs Arduino, for example).

3.14 sec :D

I use both at the same time in a special application. The esp32 manage the network and the raspi manage the analog signals and the business logic. We have two separated teams of developers, one for each uc.

you can operate up to 240MHz with the pico so in that case it could be as fast as ESP32 in my opinion

Was the WIFI module turned off on the ESP during the current consumption test?

thonny ide (uses an interpreter) made my pico run 10x slower than my nano, idk it’s useless to have 8x times 2cores more processing power and be 10x slower than a poor nano. I’ve just switched to arduino ide and it now runs about 500 times faster but there is another problem. compilation takes several minutes makes it very difficult to debug.

Note that Pi Pico can easily run at twice the frequency

Thank you

не честный тест, у 32s провода лучше от этого он быстрее

Not sure it makes a lot of sense to compare on that task, and inside CircuitPython... except if you plan to use it with CircuitPython.
I think the Pico has no floating point math and is using library to do the math... and that might not be the case for the ESP32S2 so on some mathematical activities.
I think the size of the community, the documentation, the support, ... are big factor for the maker community.
Price and availability are also important.

consider python more heavy, c++ is faster

This is a good approximation and the results are kind of expected, but it would be good to see the same test running natively instead of through CircuitPython, and without intermediate progress being sent to the display. You're spending time updating the screen and the execution benchmark is skewed by potential differences in the way the two boards communicate with the display. There's no good way to tell what these differences are here. Ideally you'd have them boot, start a timer, run the benchmark, stop the timer, and only use the display to show the results at the end.

hi, can I connect raspberry pi pico with stm32 black pill? Is it possible?

Interesting and thanks but I think it is a little lopsided as the Pi Pico has two cores and you only used one.

I’ve been playing with both. I think the ESP32 is the best option! Try to find a Pico W for less than the ESP32. Try to find a basic Pico for under the same price of the ESP32. Cost and functionality; the ESP32 wins! 🎉

Great video, exactly what I wanted to know. Thanks.

ευχαρισουμαι

awsome

Do you have video comparing esp32 and esp32-s3??

Now there is a pico W with wifi but still ESP32 boards can accomodate an external antenna, which makes them much more useful in many cases.

LILYGO® TTGO T-CAN485 ESP32 CAN Modbus rtu RS-485 Supports TF Card WIFI Bluetooth Wireless is a great project board. thanks. great video.😎

Is there anyway to do this same benchmark using both cores instead of one from the RPi?

Congrats, you proved c/c++ is faster than micropython. Since there's a C/C++ sdk for the RP2040, why didn't you run the same code on both chips?

Ok now run C on the pico...

You compare the speed of python , not the speed of the MCUs

do you have the ESP32 S3?

Second thread is very powerful, can run second task that normally would slow down your code significant on the second core, and it's easy to use in the code. For example, you can run TFT driver on the second core, or serial communication, etc. So your program runs full speed, that can be huge difference in many cases.

The ESP32 S2 takes approximately Pi seconds or 3.14 seconds and the Raspberry Pi takes approximately 2 Pi seconds or 6.4 seconds. Interesting. and a bargain price for both.

The processor chip on the pico actually has no Flash, and the Flash device is a serial Flash. This means that the chip has to load code chunks less than 264k and run them then swap them out for new code bits. This is an issue only on larger programs of course, but if you set aside large buffers for USB or WiFi processing then you could see a considerable performance hit. My guess is that the pico is between 80% slower and maybe as much as 1000% slower depending on the application, with smaller applications being favoured. It would be nice to have someone test this with a test suite designed to specifically disadvantage the pico's architecture.

Muito bom o vídeo. Parabéns.

WTF???? A "benchmark" using Python that is a interpreted language ???? Hahahahah .... Test without any foundation and that only serves to misinform and deceive the most ignorant!!
- The RP2040 (and PICO) can work at 240MHz!
- PICO W has WIFI and BT!
- RAM size and Flash size differences have no consequences! If youre writing code for a MCU with more than 2 Mbytes of flash then you are totally out of your league.
- PICO doesn't need EEPROM as programs can write directly to flash if they need to save data, there is even a file system for that. And besides... ESP with only 512 bytes of EEPROM is ridiculous and totally useless.
- PICO exposes 27 GPIO... ESP32 has only 25 usable GPIO since it uses 7 IOS for Flash Memory!
- PICO has PIO! This is a big plus! Of course, for kids and curious people who don't really know how to program and just cut and paste pieces of "arduino" code, they don't even know what PIOs are...
- PICO does not need a programmer! It has mass storage emulation, just copy the .elf file and that's it! And if we are programming in Python, we can run the program instantly from the editor/IDE without having to blink
- The PICO W costs 6€ the ESP32 costs 24€ !!
- Most importantly, PICO is all European technology and manufacturing... ESP is Chinese, not a penny more for those guys!

ESP32 is growing faster and it's getting cheaper. I would use esp32.

pi pico can run up to 250Mhz even to 400Mhz try to run it on 250Mhz and see it with same code

Whoa.... A benchmark but in Python? Not only is Python slow but the runtime for the language is probably compiled a bit different for each platform. I think straight C or C++ would be a better indicator.

This benchmark is great idea for tiny boards.

It would be nice to do a long test and include the 2nd core of the pico.
It would like to see if having 2 cores makes up for the slower proformance and if the power differences.

Can you run a simple OC on the Pico? Out of the box it seems to be underclocked, with tons of headroom.

You should compare these to a Teeny 4.1. Single core, but at 600 Mhz, which can be overclocked. Much more powerful, with a lot more I/O. I would use C libraries for the comparison however.

I wonder what result you get if you compare to the ESP32-C3 RISC V board. According to my experience, the C3 @160MHz is twice as fast as a Pico that is overclocked to 240MHz. Of course, only single core used.

Great video! Are you Spanish?

Hi, what is that display? The link is death.

Another invention of the bicycle