Raspberry Pi 5 vs. Arduino Portenta H7 vs. ESP32-S3

The Raspberry Pi, Arduino boards (specifically the Portenta), and the ESP32-S3 each serve distinct purposes, but how do they compare when put side by side? Read on for their strengths, weaknesses, and ideal use cases.

Raspberry Pi 5: A Minicomputer

The Raspberry Pi 5 is the latest iteration of the Raspberry Pi series, bringing significant performance upgrades over its predecessor. Unlike microcontrollers, the Pi 5 is a full-fledged single-board computer (SBC), making it ideal for applications requiring multitasking, networking, and graphical processing. Following is the list of the important tech specs for this device:

• Processing Power: Quad-core ARM Cortex-A76 at 2.4 GHz, capable of running full Linux distributions.

• Memory & Storage: Up to 8GB LPDDR4X RAM, microSD slot, and NVMe SSD support for high-speed storage.

• Connectivity: Gigabit Ethernet, Wi-Fi 6, Bluetooth 5.0, and PCIe expansion.

• GPIO & Expansion: 40-pin header for interfacing with sensors, motors, and peripherals.

• Power Consumption: ~3.5W (idle) to ~8W (under load).

Arduino Portenta H7: A Microcontroller for Real-Time Control

Arduino boards are known for their simplicity, but the Portenta H7 takes things to the next level. Unlike traditional Arduino boards, the H7 is designed for industrial applications, featuring dual-core processing and advanced connectivity. Key specs are:

• Processing Power: Dual-core ARM Cortex-M7 (480 MHz) and Cortex-M4 (240 MHz), enabling parallel execution.

• Memory & Storage: 8MB SDRAM, 16MB Flash, external SD card support.

• Connectivity: Wi-Fi, Bluetooth, and LTE (via external modules).

• GPIO & Expansion: 80 pins with ADC, DAC, PWM, and industrial-grade peripherals.

• Power Consumption: ~0.5W to ~1W.

ESP32-S3: The IoT Specialist

I’ll admit, the ESP32 is the board I’m least familiar with and, in fact, I have never used one although of course their existence is far from low-key, if you’re a maker, you heard about these controllers. I have to say that the balance between cost and features really hits many important points.

The ESP32 lineup is known for its wireless capabilities, and the ESP32-S3 is optimized for AI and IoT applications. With built-in Wi-Fi 6 and Bluetooth 5.0, it’s perfect for connected devices.

• Processing Power: Dual-core Xtensa LX7 at 240 MHz.

• Memory & Storage: 512KB SRAM, 16MB Flash, external SD card support.

• Connectivity: Wi-Fi 6 (802.11ax), Bluetooth 5.0.

• GPIO & Expansion: 44 programmable pins with ADC, DAC, PWM, and touch capabilities.

• Power Consumption: ~0.1W (deep sleep) to ~0.9W (active).

Technical Comparison

To make it easier for you and for me, here’s a more detailed side-by-side comparison designed to assist you in selecting the right board to use on your next project:

Final thoughts

It can be a little weird to put the Pi and the other two boards on the same comparison article - the Pi is a computer; the others are controller boards. The fact is that, at times, their features may overlap. I know that I am more experienced on the Arduino an on the Pi and I may choose these over the ESP so I can make use of that experience - makes sense. But now I know a little more and I am more likely to pick up an ESP for a new project and get out of the limitation that experience brings. Maybe this will help you explore other things too, let us know if you do!

What are the odds?

I was browsing for technical stuff as a nerd does and somehow this showed up. I immediately got distracted and started thinking whether this can be true or not. I know you’re wondering too, so let me help you (I will use Euro-millions as an example as I’m from Europe, but you can adjust this to your own country lottery).

EuroMillions requires players to pick 5 main numbers from 50 and 2 Lucky Stars from 12. The total number of possible combinations is calculated using the combination formula:

• For the main numbers: C(50,5) = 2,118,760 (see how I have used the formula here). This means there are 2,118,760 ways to select the main numbers.

• For the 2 stars: C(12, 2) = 66 (see how I have used the formula here). So, there are 66 possible choices for the lucky stars.

Since the selection of main numbers and Lucky Stars are independent events, we multiply the results together to find the total number of unique tickets: (2,118,760 * 66 = 139,838,160) This is the total number of different combinations a EuroMillions ticket can have.

For you to win the Jackpot, you must be the only one with the winning combination of numbers and stars. Therefore, if you purchase a single ticket, your probability of winning is: 1 / 139,838,160. That is approximately 0.00000072%, or 1 in 139 million.

Now we need make an assumption in order to calculate the likely hood of you going to a better place (!) on your way to get the winning ticket. Let’s say you are lazy, and you drive. Another assumption we need to make is in what country do you live? Let’s assume you are a proud Brit.

Approximately 1,695 fatalities occurred on UK roads in 2023. With a population of around 67 million, this gives an approximate annual risk of going bye-bye given by the reciprocal of 1,695 / 67,000,000:

• 1 / (1,695 / 67,000,000) =~39,500.

Statistically, for every 39,500 people in the UK, one is likely to die on a car crash annually. Puttin it all together:

• Probability of winning the EuroMillions jackpot: 0.00000072%.

• Probability of dying in car accident on the way to buy the ticket: 1 / 39,500 × 100 = 0.0025%.

Ooops.

Generate CAD from text prompts

Text-to-CAD is an open-source prompt interface for generating CAD files through text prompts. Generate models that you can import into the CAD program of your choice. Does that sound too good to be true? It does to me! I just think I’d spend more time fixing it then making the model from scratch. You be the judge: https://zoo.dev/text-to-cad

That’s all folks!

Contribute

Help this newsletter grow and improve:

• Share with friends and ask them to subscribe - I can’t wait until we’re more than the free tier of Beehiive allows for - then I’ll be asking for money 😯😯 (wait, maybe I’ll get some companies to pay for this…).

• Send me interesting topics, tutorials, projects or anything that you think may be of interest to makers and tech enthusiasts like us - if it interests you, it’ll probably interest us (send your stuff here: [email protected]).

Please do contribute and send this out to your friends. Be available, don’t be like Deepseek.