Posted by furkansahin 4 days ago
https://www.digikey.com/en/products/detail/qualcomm/QRB-2210...
$20 for something that can compete with a Pi 4 is intriguing, more so if it has a real low-power sleep state like the Pis don't. It's a gnarly chip though: 0.4mm pitch and like two dozen power rails, plus the fanout looks tight even on an eight layer board. I don't see the PMIC they're using anywhere online either... fingers crossed anyway.
Actually, now I want to build this. Maybe if these Qualcomm chips are cheap enough.
Vertical, and a bit longer than 720p? It's probably some standard size in some industry or type of device, but caught me off guard...
Historically QC chips require QC PMICs and those are usually a profit center for QC
Qualcomm being as hostile to open-source as Broadcom is definitely something in common for both SBCs.
The Fairphone 5 used a QCM6490, a qualcomm industrial chip with a 10y support timeline
The Fairphone 6 uses a Snapdragon 7s Gen 3 with ane stimated 3-4 years of support
At least for the 5 the compromise was motivated, in the second case i just don't feel like the low performance is a worthwhile compromise.
Turns out there isn't adoption success without compromises.
I think the biggest impact to hobbyists will be cheaper clone boards made with this chip.
Why was the Arduino Yun cool? It allowed me to use WiFi easily in an embedded project (this was pre-ESP32). It also allowed me to interface with a native USB device.
The Arduino Uno Q has failed on multiple fronts (in my opinion):
* You cannot leverage the enormous number of Linux USB drivers for an embedded project. It needs two USB (A) ports for interfacing with interesting devices on the SBC. A single USB-C port for power, display and peripherals is crazy.
* It makes a large point about having a GUI, but does not offer a HDMI port. Why would you do that? On the HDMI output you would have a bootloader in ROM that tests the system is okay and tells you if it cannot boot. ARM systems have an awful feature of just not booting at all and providing no feedback. "No valid OS detected in main flash storage."
* No micro-SD to flash the SBC seems like a large oversight. You want to encourage people to experiment and break the OS, and have it be super simple to restore or experiment.
* 50 euros [2] is too much for what this is. It probably should have been a single-side surface mount
* If you're going to make the HDMI output a large part of your project, do away with the LED matrix. They should be asking themselves "does this add value to all users, or could it be a shield?".
Pretty clear to me that they're aiming at drones, 3D printers and robots with that.
TIs latest version of it (AM625x) has four A53 cores, one PRU, and a Cortex M4F. It is similar in performance to the Qualcomm chip in my own benchmarks
- UI/database on desktop with multiple real-time processors connected via serial/Ethernet
- UI/db on Linux SoM with embedded processor connected through serial
- UI/db on Linux SoM and the SoM has an embedded ARM processor connected via shared memory
- UI on one core of an ESP32 with real time control on a separate core. This isn't something I've done, but many 3D printers and other low-cost machines are using this architecture right now.
I remember reading about this in some blogs so here is one I found that describes just that: https://telmomoya.blogspot.com/2016/10/asymmetric-multi-proc...
Shipping only Debian to start is fine by me. It has to start somewhere. And they seem quite responsive to making it work with other things. James Harton is plugging away at getting it working with Nerves (https://nerves-project.org) and he has it running with Buildroot already. Current repo: https://github.com/jimsynz/buildroot
Most recently they pushed their special sauce for the bootloader and how to produce the relevant mystery binaries. https://forum.arduino.cc/t/buildroot-support-for-uno-q/14108...
I share the sentiment that I don't trust that there won't be issues with Qualcomm over time. That company does have some pretty relevant chips though so I'm hopeful this means that we see them become more accessible on SBCs and embedded boards. I feel like they've been popping up more and more.
If they value this investment in Arduino they should now have a small wing of the company that pushes for things to be more open and even if they only consider that a marketing vector, if things are opened up for that purpose, quite possibly a win. But Arduino might also be absorbed into the amorphous megablob and this is the last we see. I hope not.
I don't think this board is that weird. It is just coming from the Arduino side and moving into Raspberry Pi territory. Personally I want to run Nerves on the application processor and get some practice with Zephyr on the MCU. Seems to already be supported: https://docs.zephyrproject.org/latest/boards/arduino/uno_q/d...
Also why no mention of the LED matrix. This is something RPi devices fail at. Providing som default way of neat output. First time plugging this in it starts doing fun stuff.
Qualcomm certainly seems to be saying that they want to be a different company. That they want market-share among people building products. The example elsewhere of the e-scooters using RPi's seems like the market space Qualcomm is striving to open up.
Your middle paragraphs capture a lot of the sentiment. Qualcomm is a hard company to trust. There have been a lot of neat weird interesting things that have gotten mainlined, and it's cool to see, but most products are incredibly hard to develop for, push you into vendor Board Support Packages, and don't have docs available. This chip similarly lacks technical docs.
But it sure is exciting to think maybe Qualcomm might actually want embedded market share beyond the high end of phones, routers, and laptops. And if they do want this market share, they're going to have to change.
Good luck getting your hands on any docs on how to use any of that shit tho.
Plenty of use cases in embedded!
I'm curious what the Maker community comes up with, provided it sticks around and remains supported, unlike the Edison.
I don't know how complex your Windows code is, but it's an interesting architecture to study. I bought one of the boards (they're all over Amazon) since it had everything I needed for a motion-control project, but I ended up with a simpler solution.
If you know how to eliminate the PC, why not replace it already with a Pi?
[1] https://www.seeedstudio.com/ODYSSEY-X86-v2-board-p-5075.html
Well at the moment you have to power, house and protect an entire Windows PC...
Assuming a Raspberry Pi 5 is powerful enough to run the control software, I'd go designing a carrier board for a Compute Module.
In the case of this specific product, the customer supplies their own PC and we supply the software for it. No matter how you cut the mustard, it's not an ideal setup in my eyes. It's rare to see them us an existing PC that serves more than one purpose on these machines. What I usually see them do is buy a cheap Windows laptop and have that running next to the machine. So, as I mentioned, if I can eliminate the need for a separate PC at all, that's a step forward.
I'm currently exploring the Pi 5 idea, too. I've only been with the company for a few years, but this design was made about 15 years ago? The speed of getting them to change anything, especially when what they've been doing has been working, is glacial, at best, even though they're a small company with the potential to be pretty agile. I have a Pi 5 in-house right now with Windows 10 on it (courtesy of https://github.com/Botspot/bvm) as a sort of proof-of-concept that we would not even have to port our software to work on a different platform, but it's in the corner of my workroom collecting dust due to their lack of interest. Some days, I'm not sure why I try.
I see another commenter pointed out some Seeed boards that was not aware of, so I guess it's not that the options never existed, but more so the higher-ups aren't all that interested in changing something for the sake of future-proofing or being more efficient and cost-effective. This is, unfortunately, an extremely common problem in industrial automation. That and, as much as I try to keep up, I'm also admittedly not fully cognizant of all the SBC/microcontroller options out there, so I appreciate discussions like this.
At my last job, we kept running into this problem with one customer. Every time they placed an order for the hardware we built for them, it kept getting more and more expensive because we had to search for obsolete parts and charge them a premium. But the amount that our price increased was dwarfed by the amount of money that they made selling their machine, so they literally didn't care. And as long as they happily paid us, we didn't really care that much either.