Arduino Uno Q

And here is an official quick start guide:

You can read about my own experiences of using the Arduino Uno Q in the following sections.

Please note that the images on this page are available for download from the bottom of the page for better viewing.

Getting started

In March 2026 I bought an Arduino Uno Q 4GB RAM 32GB eMMC version, model number ABX00173, from Farnell for £54.05 ($71.89).

The Uno Q arrived boxed, on the top of the box is an illustration of the board, on one side of the box is the model number ABX00173, on the other side URL arduino.cc/uno-q and a QR code (which I couldn't get to work). On the back of the box it has written:

UNO Q: WHERE ARDUINO SIMPLICITY MEETS THE POWER OF AI, LINUX, AND REAL-TIME.

Connectivity Wi-Fi Bluetooth LE

MPU Qualcomm Dragonwing QRB2210

MCU STM32U585

Memory 4 GB LPDDR4 32GB eMMC

Interfaces USB Camera 2x MIPI CSI Camera Monitor Modulino Keyboard Mouse

There's also the same URL and QR code as on the side.

On another side of the box is a different QR code, which does scan but is simply the text ‘T03H85’ (which is written under the QR code) and barcode 7630049205727.

Inside the box is the Arduino Uno Q in an anti-static bag, along with a sticker showing a QR code and the same URL as before (I did get the QR code to work, it points to the same URL), and a booklet with general information (warranty, manufacturing, testing, etc.). Note that the booklet mentions model number ABX00162 which is the 2GB RAM 16GB eMMC version. It has the link for the user manual: docs.arduino.cc/manuals/ABX00162, which is for both variants.

I’ll mention some components of interest (the lower spec Uno Q will have some differences):

PCB top

POWER button.

USB-C female.

JCTL 2x5 male header.

Arduino Uno female headers; 1x18 along one edge, 1x14 along the other, with pinout on both side of both headers. Unlike older Uno, which had 4 female headers, Uno Q only has 2 headers, with a blanked pin to separate what was originally 2 headers for both sides. This means there isn't the issue of incorrect header spacing like with older Uno.

SlimPort ANX7625 IC.

27MHz crystal.

WCBN3536A module.

5TB77 D8CJC IC.

Qualcomm QRB2210 IC.

Qualcomm PM4125.

T384 MLH1 crystal.

RGB LEDs x 4. LEDs ‘1’ and ‘2’, which are controlled by the MPU, hence QRB written above, and LEDs ‘3’ and ‘4’, controlled by MCU, hence STM written above.

JSPI 2x3 male headers.

QWIIC connector.

13 x 8 LED matrix.

The PCB features the Arduino logo along with the text ‘UNO Q’.

PCB bottom

ARDUINO.CC is written on the PCB with 7 small gold circular pads above it.

Two 2x30 female high density connectors, one labelled ‘JMISC 1.8V’ & ‘JMEDIA’.

KLMBG1UCTC IC.

H165L crystal.

STM32U585AII6 IC.

QR code (I couldn't get it to scan) and model number ABX00173.

Arduino and Dragonwing logos printed on the PCB.

2x3 small gold circular pads.

Written on the PCB is Arduino S.r.l. Address, serial number, FCC ID 2AN9S-ABX00162 and IC 26792-ABX00162. I couldn't find any detailed reports from searching the FCC ID, however from the IC number I found:

https://sms-sgs.ic.gc.ca/equipmentSearch/searchRadioEquipments?execution=e1s11&index=0

Model number ABX00162, with approval date of Fri Dec 26 06:30:47 EST 2025

https://sms-sgs.ic.gc.ca/equipmentSearch/searchRadioEquipments?execution=e1s12&index=1

Model number ABX00162, with approval date of Mon Oct 20 00:00:00 EDT 2025

https://sms-sgs.ic.gc.ca/equipmentSearch/searchRadioEquipments?execution=e1s15&index=2

Model number ABX00173, with approval date of Fri Dec 26 06:30:47 EST 2025

For testing the Uno Q I followed the user manual:

https://docs.arduino.cc/tutorials/uno-q/user-manual/

As mentioned, you need to use Arduino App Lab 0.1.23+, only the MCU part of Uno Q can be programmed with recent Arduino IDE version.

The page has UNO Q Board Architecture Overview, pinout, datasheet, schematic, and STEP files.

For powering UNO Q can use a USB-C cable and power supply providing 5V DC 3A,+5V DC power supply connected to 5V pin, or +7-24V DC power supply connected to VIN pin.

The image mentions USB-C requests 5V power delivery profile and powering through 5V pin requires 3A but no current requirement mentioned for VIN pin as likely depends on the voltage supplied. There is a link to a page with much more information about power:

https://docs.arduino.cc/tutorials/uno-q/power-specification/

Arduino App Lab

I installed on my PC Arduino App Lab (V0.6.0), which is designed for high-performance computing boards.

As well as running Arduino App Lab on PC connected to the Uno Q, it can also be run on Uno Q in standalone mode (4GB RAM variant recommended); Arduino App Lab comes pre-installed on the Uno Q.

To use Uno Q in standalone mode you need to use a powered USB-C hub connected to Uno Q, with monitor, keyboard, mouse and power connected to the hub. Please see the Standalone mode section for more information.

You can also access the UNO Q from any PC wirelessly on your local network using Network Mode.

To install Arduino App Lab on PC see:

https://www.arduino.cc/en/software/#app-lab-section

I opened Arduino App Lab, as you can see below:

This was without the board connected.

I then connected the board to my PC, the green power LED came on, the LED matrix showed an animated Arduino logo then an animated heart graphic, the blue ‘2’ LED came on then also the green ‘3’ LED was flashing.

At this point the board was detected:

In Device Manager I noticed ‘USB Serial Device’ under ‘Ports (COM & LPT)’, and ‘ADB Interface’ under ‘Universal Serial Bus devices’.

In Arduino App Lab I clicked on the ‘uno-q’ board image and it took me to the configuration page:

I selected the keyboard layout, gave the board a name (cannot be the default 'Uno-Q'), and connected to a Wi-Fi network. Once connected it looked for updates and found more than a few. During the updates the board’s blue ‘2’ LED was flashing, the updates took ~11 minutes.

While the updates were necessary to ensure smooth running of the board and there were a lot since it hadn't been updated before, it's certainly not as quick to get started with Uno Q compared to the traditional Arduino boards, the price to pay for the extra processing power and complexity of the Uno Q.

After updating, it prompted to restart the Arduino App Lab, I clicked the button and the Arduino App Lab showed my board (which I named Uno-QJS), accessible both via USB and on the network:

I selected the board using the USB option, the app said it would check for updates but then nothing happened. I clicked several times more but no response.

I tried clicking on the network option and it asked for the password, which I hadn't set up, apparently the default password is arduino, however, it wouldn't let me enter the password - it could have been a wireless issue.

I closed down the Arduino App Lab and reopened it but I had the same issue when I tried selecting the board, both through USB and wireless.

I disconnected the board from the PC, plugged it back in, this time it didn't do the animations on the LED matrix. I reopened Arduino App Lab as administrator and the board showed up as if it were the first time setting it up. It looked for updates and then moved to a page asking for the Linux credentials (see below) as part of the original setup, so it looks like after doing the updates it skipped ahead without finishing the setup. Looking online, one suggestion for a similar issue was needing to set the Linux password, so that makes sense.

I kept the username as the default 'arduino' and entered a password.

A Welcome to Arduino App Lab screen showed:

Which explains about apps, bricks and examples.

It then went to the examples page:

There are examples from simple to complex, and along the left side are various options; My Apps, Bricks (see below), Learn, Settings, and Account.

Clicking on Settings it has some board information including the FQBN unique identifier (arduino:zephyr:unoq), and disk storage showing (4.50 GB of 27.62 GB used). Hovering over the ‘i’ symbol it gave further storage information:

ROOT 3.92 / 9.72 GB

USER 0.59 / 17.89 GB

There were also options to change the password, and reinstall the O/S.

Near the bottom of Arduino App Lab it shows the connected board, along with a button to connect to the board’s shell, and a ‘System stats’ which clicked on reveals:

STORAGE : ROOT 3.92/9.72GB

USER 0.59/17.89GB

RAM: 0.32/3.58GB

CPU: 3%

I clicked the ‘Blink LED’ example which opened it up:

It auto selected README.md, which explains how the example works and how to run it. We can see it doesn't use a brick but does use the router bridge, Python and Arduino sketch code. This shows how even to do something simple it's much more complicated than doing  thesame on a traditional arduino.

Along the left side it shows the sketch libraries (Arduino_RouterBridge), and various other files, in particular main.py under python and sketch.ino under sketch, which I show below:

To run the app the Run button top-right has to be clicked. If select the drop-down next to the Run button there is radio button to set the program as a startup app.

When I clicked on the Run button, the ‘3’ red LED came on, and after almost 7 minutes the example was running: the ‘3’ red LED was flashing. The long time it took to run the sketch shows the big disadvantage of doing something simple on a complex board.

Looking at the information that is reported in the console tab, the example sketch used 21078 bytes (2%) of program storage space, maximum of 786432 bytes. Global variables 5036 bytes (3%) of dynamic memory, leaving 126036 bytes for local variables. Maximum is 131072 bytes.

You can click the Stop button bottom-right to stop the sketch running, the ‘3’ LED will go off and the ‘2’ LED will come on.

Apparently anti virus software can slow down Arduino compilation so I uninstalled Norton 360 (which I was going to remove anyway as it was preinstalled on my new PC). I restarted the computer, opened Arduino App Lab and connected the Uno Q to my PC. This time I timed it:

Upon connecting the Uno Q to my PC the Arduino logo displayed on the LED matrix followed by the heart animation. It took ~30s from plugging in the Uno Q to my PC for the LED matrix to become blank and in total ~1m 10s for the board to be detected by the Arduino App Lab. When I clicked on the board in App Lab it asked for the Linux password, after entering it and waiting ~5s Arduino App Lab showed the examples.

I clicked on the Blink LED example again, clicked on Run, this time it took ~12s, which is a vast improvement over the first time but whether it had cached the compilation I don't know. So I clicked on the Stop button and then Copy and edit app button to the the right of the Run button. I entered a name, and clicked the ‘Create new’ button.

Clicking on main.py I changed time.sleep(1) to time.sleep(5), so there would be a noticeable difference.

Note how both Python and sketch (C/C++) files are required just to do a simple blinking of an LED. I then clicked on the Run button and it took ~50s to run the sketch so indeed perhaps Norton was slowing it down when i first ran the Blink example.

I then set the sketch to run at startup by clicking the drop-down to the right of the Run/Stop button and toggled on ‘Run at startup’. Note that the built-in examples can't be made to run at startup, only a copy of an example or new app can.

The copy of the Blink example now appeared under My Apps and because I had it set to run at startup it was marked as default. I’ll check next time I power cycle the board if it does actually run the app.

Running AI was something I was especially interested with the Uno Q so I tried out the ‘Bedtime story teller’ example, which creates a bedtime story using a cloud-based language model to generate the story based on a user interface through a web interface.

Running the app is a bit involved. First you need to make a copy, since you'll have to edit the app. After making a copy, you have to click on ‘Cloud LLM’ under Bricks on the left side and click on the ‘Brick configuration’ button on the right side. You will then need to enter an AI API key, links are provided in the readme. Note that this may incur a cost, check with the provider. After entering the API key, click the Save button.

The readme explains main.py handles the AI and data flow (there is no sketch file). In the ‘Understanding the Code’ section it lists the three Ai providers (Google Gemini, OpenAI GPT, and Anthropic Claude) that can be used. Near the start of main.py CloudLLM() is called with a particular AI model, which can be changed. By default Google Gemini is used, since I had created a Google API key for this example I didn't need to change anything. So all I needed to do was click the Run button, which gave me the warning that I was about to change the app for the run, as I still had the modified Blink example running. I clicked the ‘Confirm and replace’ button. It took only about 9s to run (probably because it was just using Python) and automatically opened the browser with the Bedtime Story Teller page opened. The readme does have the link in case you need it.

Note that Wi-Fi speed and AI service speed in addition to the Uno Q’s processing will affect how quickly the AI responds.

As can be seen below, there are parameters to customise the story or you can click the ‘Generate Randomly’ button, which is what I selected. However I got error ‘An error occurred while generating the story: Response generation failed: No generation chunks were returned’. I checked my Google account and an API request had been made but no error reported.

Wanting to try again, I clicked on the ‘New story’ button, confirmed that I wanted to create a new story and clicked the ‘Generate Randomly’ button again. This time it was successful, generating a story after ~19s.

The story was written adequately, consisting of an introduction, 3 short episodes, and an epilogue.

I created a new story, this time setting the parameters, which included ‘Long 20+ min’. It took ~20s but the story was only slightly longer than the previous one.

While it's quite impressive to run AI on such a small and relatively cheap board the real potential is interfacing AI with lights, sounds, etc. to interact with the real world.

The next example I tried was ‘Blink LED with UI’. It uses Linux to change the LED state using basic event handling and UI updates. It took ~57s to run and automatically opened the browser page consisting of a button that lets you toggle the Uno Q’s ‘3’ LED on/off, which changes instantly. A very simple example but it shows the basics of UI control.

Next I tried Cloud AI Assistant, a simple chatbot using Cloud LLM. As with the ‘Bedtime story teller’ example you need to make a copy of the app, add the AI API key, and edit main.py if you want to use an AI service other than Google Gemini. It took ~8s to run and it automatically opened the browser page.

There were a number of preset prompts along with a box to enter your own prompt. I entered prompt “What is the Arduino Uno-Q?” and got a response in ~5s, however, it claimed to not know what an Arduino Uno-Q is. Asking the same thing to Google Gemini via a browser on my PC it did know what an Arduino Uno-Q is.

Returning to Uno-Q, I reset the chat using the provided button and asked the same question but it still insisted to not know about the Arduino Uno-Q. I’m not sure where the fault is.

I tried a different prompt, asking what a PIC microcontroller is and it gave a satisfactory answer. Interestingly, it said PIC stands for Programmable Intelligent Computer instead of Peripheral Interface Controller, the original name, although PIC nowadays doesn't really stand for anything. Note that Google Gemini from the browser on my PC specified ‘Peripheral Interface Controller’.

I did do some other prompts and it got some details wrong that Gemini didn't when run on PC.

Next I tried ‘Weather forecast on led matrix’, which uses the open-meteo.com service to show real-time weather information on Uno Q’s LED matrix, using Python for API and weather processing, and Arduino sketch for handling the LED matrix animations, updating every 10s. Before running the app you need to make a copy and edit sketch.ino to set the city variable as required.

It took about ~58s to run and an animated rain cloud appeared on Uno Q’s LED matrix, which didn’t reflect the weather in the city I had entered (London), although a small chance of rain was forecast. In the Python tab in Arduino App Lab it reported that drizzle was forecast for London.

Oddly, the LED matrix animation froze after a few minutes. I stopped the app and ran it again and the animated rain cloud appeared again only to stop again after 2 minutes. Since the Python tab was no longer showing the current weather report it made sense the animation had stopped but I don't know why.

Next, I tried Uno Q Pin Toggle, which lets you change the state of the Uno Q pins using a web interface. It uses Python for handling the web interface and a sketch to set up and control the I/O. It took ~58s to run and ti opened up a browser page with illustration of the Uno Q and radio buttons to set the pin states.

Starting off simple, the on-board ‘3’ and ‘4’ LEDs can be controlled, the button toggles are toward the bottom of the page. To test the other I/O you will need to connect an LED with series limiting resistor (I used 1K) between the pin and GND. I started with D0 but the LED didn't light when I toggled the pin on, so I tried the onboard LEDs again and there was no response. Returning to Arduino App Lab I tried stopping the app with the Stop button but the Stop button didn't change state. So I closed down the Arduino App Lab, disconnected the Uno Q, reopened Arduino App Lab and reconnected the Uno Q.

Running the Uno Q Pin Toggle example again I was now able to toggle D0 on/off. Considering I had problems with the Weather forecast example, which stopped after a couple of minutes, I wonder if it's a related issue. The cable connecting Uno Q to PC is a known good cable and the PC is new.

I measured between D0 and GND when toggled on and got 3.296V to GND, which matches the 3.3V logic level the MCU runs at.

I proceeded to test a few of the other pins and didn't have any issues with the app no longer working.

Next, I tried Classify images example, performing image classification using a pre-trained neural network and a web-based interface which lets you upload images for classification, with setting for confidence threshold and to see the inference results.

Only python code is used, main.py, which has the browser interface for reading the input, decoding, and inference. It took ~3m 30s to start running. I uploaded an image of a dog, kept confidence level at the default of 0.50 and clicked the Run Classification button. It took about ~20s but I tried 2 different clear images of a dog and each time it failed to detect objects. I tried with the confidence level set to 0.25 and then at 0 but still it was unable to detect any objects. I then uploaded an image of a cup of coffee with the confidence level still at 0, this time it worked, having 91.77% confidence of it being espresso, which was right. I tried again with the same image but confidence level now at 0.5 and again it detected the object correctly with the same level of confidence.

Image size didn't seem to be issue as the coffee cup image was high resolution and so was the dog image.

When trying images of other types of objects it either didn't detect the object or got it wrong (example, image of upper half of a woman detected as a wig with a confidence of 92.20%).

Annoyingly it wouldn't let me stop the running app so I had to unplug the board, close and reopen Arduino App Lab AND plug the Uno Q back in.

Another time I ran the Classify images example again and it took ~24s to start, the first time it had to download a lot so perhaps thats why it took a lot longer to run.

Next I tried ‘Detect objects on images’ example. Whereas the Classify images example tries to put the object(s) in the image into a class, this example tries to label individual objects.

The ‘Detect objects on images’ example uses a pre-trained machine learning model, demonstrating how to process input images, run inference, and visualize detected objects with bounding boxes and labels. The example uses python, main.py, which handles the input image, decoding, inference, drawing the bounding boxes and results. The example took ~22s to run and has the same browser page interface as the Classify images example.

It took ~55s to run the detection after choosing an image and it correctly detected it was a person with 88.9% confidence (see image below). Considering the woman wasn't face on it was quite impressive that it correctly detected the person.

I tried other images, including those that weren't successful with the Classify images example and it successfully detected the objects, including multiple objects when there were more than one in the image.

The larger the image size the longer the detection took.

Next i tried Mascot Jump Game (see the following image), an endless runner game controlling an LED character jumping over electronic components, with progressively increasing difficulty, score tracking, one-button gameplay, and synchronized LED matrix animations on the Uno Q.

Space/up arrow is used for jumping, R to restart after game over.

The python file, main.py, manages game state and physics calculations, the sketch file handles the LED matrix animations.

The game plays in the browser (remember this is running and hosted on the Uno Q), and the Uno Q’s LED matrix shows the game’s state. Apart from infrequent slowdown the game ran smoothly.

What I have not found a clear answer to is how to shut down the Uno Q, even the official manual doesn't cover it and there is conflicting information from other sites. Unlike a traditional Arduino that you can just unplug from the power source (assuming it's not writing to EEPROM/flash memory), because the Uno Q runs Linux you should safely shut it down to avoid corrupting files. Please see the Tips/Shutting down section for solutions.

I tested current draw with the Uno Q connected to a PC using a USB volt/amp meter, while the Uno Q was booting the peak current draw was 0.56A. Once the board was detected by Arduino App Lab the current draw fluctuated between 0.12A to 0.17A. During compiling of an example I saw peak current draw of 0.75A. So I don't know why the Uno Q needs 3A when run standalone as connected USB peripherals shouldn't draw that much but perhaps it's to allow for multiple devices that draw high current.

Some further notes:

Sometimes, when connecting the Uno Q to my PC after opening Arduino App Lab, or connecting the Uno Q to a PC and then opening Arduino App Lab, it will ask for the Uno Q’s linux password.

I noticed that the Arduino App Lab started showing an IP address at the bottom and when I opened up the shell it displayed:

The authenticity of host '192.168.1.247 (192.168.1.247)' can't be established.

ED25519 key fingerprint is SHA256:YTvroYZcK5TGTgR4cxWnkdhnI3lHVhZGFbqiYYIQMLE.

This key is not known by any other names.

Are you sure you want to continue connecting (yes/no/[fingerprint])?

I entered yes, and then the password, and then I got the usual Linux password.

Standalone mode

While using the Arduino Uno Q in standalone mode you develop apps on the Uno Q using the Arduino App Lab without the need for an additional PC but as you have access to the full Linux experience on the Uno Q you can use it like a regular PC although of course the limited processing capabilities of the Uno Q won't replace your main PC.

I first tested Arduino Uno Q's standalone mode using a USB-C hub (no model number) which I had bought from Temu which supports PD, along with a Raspberry Pi compatible USB-C power supply I happened to have, model number MKE2-P27WSUKD, which can deliver 5.1VDC 5A (5V should be fine but the Raspberry Pi power supplies tend to be rated for 5.1V). I plugged the power supply into the hub’s USB-C PD port and a keyboard and mouse into the hub’s USB ports and a HDMI cable from the hub to  to my TV. I connected the hub’s integrated USB-C cable to the Uno Q’s USB-C port. Plugging the Raspberry Pi power supply into the wall socket, the Uno Q showed its boot animation on its LED matrix and after ~19s I saw boot text like on my TV like you would expect from Linux but even after waiting 3 minutes nothing else appeared on the TV (there’s supposed be a login screen). I power cycled the Uno Q and this time I looked closer at the boot text but saw nothing suspect, such as an error message. I tried a monitor instead of the TV but I still didn't get anything more than the initial boot text.

Some months later I bought from Amazon a different USB-C hub, in case the other one I tried wasn't delivering enough current to the Uno Q to get to the login screen. The new hub I bought was a UGREEN CM478 USB-C Multi-function Hub for a little over £8 ($10.58). The hub features x1 HDMI, x2 USB 2.0, x1 USB 3.0, x1 USB-C PD, and integrated USB-C cable. The PD input rated for 5.0-20.0VDC 5.0A Max.

This time I used an official Raspberry Pi USB-C power supply, model P3619, rated for 5.1VDC 3.0A, the only other USB-C power supply that should work, just in case the other power supply was at fault even though my suspicion was the Temu hub. I plugged the P3619 power supply into PD port of the new hub, HDMI from my monitor into the hub, and  the hub's integrated USB-C cable into the Uno Q. I did not use a keyboard or mouse as I just wanted to see if the Uno Q reached the login screen.

I plugged the power supply into the mains and boot text appeared on my monitor as before and after ~37s total the login screen appeared. So clearly the previous hub and possbily the power supply I had used was the problem. I plugged a keyboard and mouse into the hub, both of which worked immediately. However, I had forgotten the password I had previously set so I tried to shut down the Uno Q by clicking the power icon top-right and selecting ‘Shut Down…’ followed by clicking the ‘Shut Down’ button to confirm. But, just like when I tried to shut down the Uno Q while connected to my PC, but it actually restarted, shortly returning me to the login screen. I tried again to shut down but it again restarted so I pulled the power.

I then connected a Logitech keyboard and Raspberry Pi mouse (both items I had around) to the hub and plugged in power again. After logging in (I had found the password) it took about 15s to load into the Arduino App Lab (version 0.6.0). There was pop-up window listing the new features in the update and an option to skip or install the updates. I did the updates as it contained bug fixes as well as more features including lots more examples to use in the App Lab.

The update took 15m 21s, during which progress text was shown. I had to click the ‘Restart App Lab’ button to restart the App Lab and it took about 10s before the ‘Welcome to Arduino App Lab’ window appeared. I closed the window to reach the main App Lab interface but oddly was still showing as version 0.6.0 (App Lab was supposed to be updated to 0.10.0).

Under Settings Disk Storage was showing as ‘Used: 7.73 GB of 27.62 GB (ROOT 6.66 / 9.72 GB USER 1.07 / 17.89 GB). Also, it showed Build Version as 20251111-426, Linux Distribution as Debian GNU/Linux 13 (trixie) and Release date as 11/11/2025. Clicking On View release notes brought up a Chromium browser window with release notes showing App Lab version is 0.10.0, which was released Aug 18.

By minimizing the App Lab window you can view the Linux desktop and along the bottom of the screen are various shortcuts including Terminal Emulator, File Manager, Web Browser, and so on. To return to App Lab from the desktop simply click its name at the top of screen.

To access all applications, click Applications in the top-left corner. The apps (some of which are in their own section) include Terminal Emulator, File Manager, Web Browser, various settings apps, Ristretto Image Viewer, etc. I selected Display from Applications->Settings to check the resolution, which was set to 1920x1080 @ 60.00 Hz. By clicking on the resolution drop-down, there were options ranging from 640x480 up to 1920x1200, the latter of which had asterisk by it. 1920x1080 is supposed to be the Uno Q’s maximum resolution but changing to 1920x1200 worked without issues. 

Apps open quickly, from ~2s for simple apps to ~4s for more intensive apps, such as Chromium, for example. Using eMMC storage, rather than SD card which is typically slower, gives Uno Q a speed advantage over other single-board computers, such as a typical Raspberry Pi. 

Web pages of course don't load ultra fast, with YouTube being particularly slow for example, taking ~18s to reach the YouTube home page, using the Chromium browser. Clicking on a video it took ~14s for the video to start playing, at 480p there were 29 frames dropped out of 1000, which was pretty good. Of course this is not the primary use of the Uno Q but shows that it is quite capable of being used as a general purpose computer.

The Uno Q was connected to my router via Wi-Fi and I did a couple of speed tests, first running ping www.google.com in the terminal, which reported average of 16ms. I then did a speed test from the Google home page and that reported 76.4Mbps download/2ms latency. So you can see that any slowness with loading pages on the Uno Q was due to the site's response time and the Uno Q's limited hardware processing speed rather than primarily the Wi-Fi speed.

I found that the time displayed was an hour off so I adjusted it by right-clicking the date and time top-right of the screen and selecting Properties. In the Time zone box I entered and selected Europe/London, I also set the time format to be appropriate for the UK.

Attempting shut down (Applications->Log Out, click Shut Down button) restarted the Uno Q instead so I pulled the power when it got to the login screen.

Next time I used the Arduino Uno Q in standalone mode it showed App Lab as version 0.10.0 so perhaps when App Lab was supposedly restarted it didn’t properly restart but power cycling actually loaded the updated App Lab.

Later I tried the Raspberry Pi compatible USB-C 5.1/5A power supply (that I had previously mentioned) on the new hub and it worked, showing that the problem was indeed the Temu hub. However, when testing the new hub, I had plugged the power supply into mains first and as I plugged its USB-C cable into the hub, as the connector brushed against the hub case it sparked. This only happened when the hub was plugged into the Uno Q, so possibly the Uno Q was creating a potential difference. It doesn't seem to have damaged anything but in future I’ll plug the power supply USB-C cable into the hub first before plugging into the mains, regardless of what power supply I use.

Another time when I was using the Uno Q in standalone mode I measured the power used by the Uno Q by sticking a USB-C digital meter between the Uno Q’s USB-C socket and the hub’s USB-C cable. This type of meter alternates between showing the voltage and wattage. From when the Uno Q started booting to when App Lab appeared after logging in I saw peak power draw of 2.7W, which is to be expected going by what I found online.

Tips

Linux

Please see this link for the official Arduino Uno Q Linux guide:

https://docs.arduino.cc/tutorials/uno-q/debian-guide/

Shutting down

Since the Arduino Uno Q runs Linux it should be shut down properly like you would with any other computer and this differentiates the Uno Q from other Arduino boards. However, the Uno Q supposedly doesn't have the hardware capability to perform a shut down and indeed, trying to shut down the Uno Q will trigger a restart instead. This can be seen when selecting shut down in Linux when the Uno Q is being used in standalone mode, or by using typical shut down commands (sudo halt, sudo shutdown -h now, etc.) from the terminal either in standalone mode or from the App Lab's shell.

In May 2026 the 'Debian Linux Basics for UNO Q' guide was updated with details of how to safely shutdown in the 'Shutting Down Your UNO Q Safely' section:

https://docs.arduino.cc/tutorials/uno-q/debian-guide/

Troubleshooting

Updates

If updating the Arduino Uno Q isn't successful keep trying as looking online people had to make multiple attempts to get the Uno Q updated.

All content of this and related pages is copyright (c) James S. 2026