This Home Made Laptop Raises The Bar

With ready availability of single board computers, displays, keyboards, power packs, and other hardware, a home-made laptop is now a project within most people’s reach. Some laptop projects definitely veer towards being cyberdecks while others take a more conventional path, but we’ve rarely seen one as professional looking as [Byran Huang]’s anyon_e open source laptop. It really takes the art to the next level.

The quality is immediately apparent in the custom CNC-machined anodised aluminium case, and upon opening it up the curious user could be forgiven for thinking they had a stylish commercial machine in their hands. There’s a slimline mechanical keyboard and a glass trackpad, and that display is an OLED. In fact the whole thing had been built from scratch, and inside is an RK3588 SoC on a module sitting on a custom-designed motherboard. It required some effort for it to drive the display, a process we’ve seen cause pain to other designers, but otherwise it runs Debian. The batteries are slimline pouch cells, with a custom controller board driven by an ESP32.

This must have cost quite a bit to build, but it’s something anyone can have a go at for themselves as everything is in a GitHub repository. Purists might ask for open source silicon at its heart to make it truly open source, but considering what he’s done we’ll take this. It’s not the first high quality laptop project we’ve seen by any means, but it may be the first that wouldn’t raise any eyebrows in the boardroom. Take a look at the video below the break.

 

 

 

https://www.byran.ee/posts/creation/

54 thoughts on “This Home Made Laptop Raises The Bar

  1. CNC-machined anodised aluminium case

    nice

    slimline mechanical keyboard and a glass trackpad

    nice

    display is an OLED

    nice

    RK3588 SoC

    WTF LMAO

    Don’t waste your time on the video, in the end it’s yet another fancy case for an R-Pi clone.

    1. Keep in mind that what we have here isn’t just a product to make judgement on, but rather a project that shows how to achieve a goal, and you might want to examine it even if your goal isn’t identical. This is, after all, HackaDay.

    2. Don’t waste your time on the video, in the end it’s yet another fancy case for an R-Pi clone.

      It’s opensource, dude, what’s the problem to route your own motherboard for this case with x86 processor and some Nvidia/AMD GPU?

      Ooops, sorry, Intel/AMD and Nvidia is the most proprietary architecture in the world, so you are not allowed to really make your own motherboard like you can for Rockchip/Allwinner/etc. ARM/RISC-V/MIPS SoC’s. You are not even allowed to have your own BIOS/UEFI/u-boot for your beloved x86.

      Say thanks to Intel/AMD/Nvidia, for their love to screw up end-users.

    3. The Pinebooks use similar Rockchip SoCs. They have ARM cores. They’re plenty powerful, more open-source friendly, and not as power hungry as x86 trash. They’re an excellent choice for this project.

      1. I’ve been using my pinebook pro a fair amount over the past few days. Does especially well with a lightweight DE or WM. Browsers can be a bit slow, but for writing, coding (unless you’re trying to compile a largish project), document perusal, KiCad, and other similar workloads it works great.

        1. The Manjaro install on my Pinebook pro borked itself somehow just before xmas, I’ve not gotten around to investigating or reinstalling. I’ve only used the original linux (whatever it was) and Manjaro on it. Which OS do you run on yours?

      2. there are rockchip arm chips that are very efficient with their power consumption. rk3288 in the original asus chromebook leaps to mind.

        but — and i don’t know if this is amenable to hacking kernel drivers, since the rockchips are relatively open — there are also some that perform abysmally.

        and one more piece to the puzzle: modern intel celeron ‘n’ series chips are mostly incredible at power management. much better than raspberry pi (broadcom), for example.

        so like 10 years ago the rule of thumb was that arm anything is more efficient than intel anything. but that’s simply not true anymore. cycles vs watts, i’m not sure who would win…but there are ARMs now that are hot even when they aren’t in use! and there are intels that hardly draw any current so long as the CPU is executing HLT. and in a portable computer, that’s what really matters. can it quickly transition to and from a low-power mode?

        1. x86 trash? That trash is running this world.

          If you count ARM/MIPS/Xtensa/etc CPUs around you right now, you will be amased how small is the share of x86 today.

          It’s like to believe that some sparkling and pesky celebrity runs the world, not all that people who have real jobs and do real things.

          The worst thing that 99% of x86s do nothing really useful, running browsers, social network farms, datacenters with useless stuff and all that unnecessary bells and whistles to make all this useless stuff even more resource hungry to force you to buy new x86. It is a shame when complex CAD/CAM/Scientific/Medical/etc. software runs smoothly on decade old PC, while browser could eat all memory and CPU of a new and much more powerful x86.

          And all that tiny ARM/MIPS/etc. you despise silently doing their job around you 24×365.

    4. Somebody’s project doesn’t meet your needs, maybe they show off some skills you could develop to help you meet your needs.

      Like choosing a different SoC.
      The cpu choice is the least interesting part of this in many respects, but that’s no dis on the project.

      Now I need to finish one of several projects I’ve got on the fly. None of which are this scale, or fit and finish.

    5. There are several design decisions in the build process. While some choices aren’t what I would make, the fact remains that those choices are documented and followed through to a complete product.

      If I was to make something remotely as good, I’d feel pretty pleased about it.

  2. At first I was like, interesting, but the look was just not for me. Then I saw the SOC, and well, nope.
    I get this could be a fun little project, but I do serious work and this thing would not handle my day to day software. And please dont start on open source alternatives… they cant compete

    1. It’s not the software that can’t compete but your mind that got locked itself into Micr🤬😳😠it ecosystem 🤣

      Unless you’re doing some space shuttle stuff there’s no reason why you cannot install Linux Mint instead of that cancer written by a bunch of american nerds.

      1. Who said anything about Microsoft?

        In my experience with these kinds of modules and SBCs, the issue is often not the OS, but rather the architecture. While ARM support is getting better, I’ve still run into many situations where x86 was the only option. Sadly, small and capable x86 SOMs and SBCs seem less pervasive (though the LattePanda Mu modules seem interesting, especially the new version).

      2. Why do you care so much about the operating system other people use?
        You are not a superior being because you use Linux and people like you feed the stereotype of Linux enthusiasts being elitist jerks.

        Windows and Microsoft products in general are fine and very likely allow you to use hundreds of products and services even if you yourself never directly touch the software. Linux is free and open source so if Microsoft was not providing companies with some sort of advantage, they would not be spending money on it. It’s that simple.

        I love Linux and it powers a bunch of apps and services in my home, and in my home lab. I am always SSHed into one or two of those instances for one reason or another. My ebench laptop also runs Linux. That said, I prefer to run Windows on my main productivity machine. It’s what works for me. If you prefer a Linux desktop, that’s great too, but don’t be so obnoxious about it and respect other people’s preferences.

    2. This perspective always fascinates me because IME proprietary alternatives can’t compete with the FOSS software I use, but clearly there are a few niches dominated by corpo software. What specifically do you have in mind?

  3. Gorgeous build quality, great documentation.. especially impressive for a friggin senior project. Someone’s gonna get mad job offers!

    My little critiques:
    – ESP32-S3 seems like crazy overkill for just running power controller… but with it connecting over USB, its role could always be expanded later for more fun stuff
    – modular keyboard is sweet, but if the magnetic attachment is power only, that’s a missed opportunity…wired keyboards ftw!
    (couldn’t tell 100% from write-up)

  4. I know this is the internet and thoughtless comments are the norm but I still never understand the torrent of negativity that projects receive here. And always by people who don’t share links to their own projects so we could marvel in awe at their expertise .

    Byron did a complex thing that a generation ago would have been impossible for an individual to have done. His work spanned mechanical engineering to high-speed electronics design to software development. He did a lot of research and he made reasonable choices in selecting components such as the SOC module, OLED display and hinges. His removable keyboard is a clever idea and the overall computer very professional looking. He never intended it to be a finished product. It scratched an itch. An itch that I’m sure most of the naysayers would not have the skill or expertise to tackle.

    I understand I’m probably fruitlessly shaking my fist at the sky but is it too much to ask for some of you to actually think a moment before typing? Let’s celebrate each other instead of constantly tearing down.

    1. Agreed, mostly anyway. As

      that a generation ago would have been impossible for an individual to have done.

      that just isn’t really true, that generation ago you’d have been building your custom z80 or something, capacitive touch would likely be replaced by a trackball if you have a pointing device at all, ‘high speed signalling’ would have meant something really quite different – the world has moved on a great deal. But you can and folks did create similarly impressive personal twists back then.

        1. Hmm true enough, though Z80 and its Ilk were before my time anyway picked as a long lasting and very identifiable period of old tech. Also it and similar chips remain dominant in the ‘small’ form factor portable stuff for a prolonged period. Doesn’t really negate my point though, if you could do it two generations ago that is even better than just one…

    2. Compare the response here to that of Hacker News. Over there, this project is a masterpiece of mechanics and electronics; over here, it’s “he bought a bunch of stuff and put it in a case”. It’s not like we don’t see a “homemade laptop” once every couple months on HaD anyways. Or an “open source” computer that uses closed source hardware building blocks? A critical eye on this project will see something that well, looks like a high schooler put it together. But if this post here were about some sort of clever software thingy, the responses would be exactly reversed.

      The standout feature to me on this project is the time commitment. Byron took many, many small steps over several months to get here. No step is all that impressive by itself to any seasoned HaD reader, but maintaining the focus and drive for long enough to get to this point should really be applauded. I wish I was able to dedicate as much effort to my own projects.

      1. I saw this on HN before HaD, and I noticed the same thing. I prefer the HN positivity. I also happen to spend more time there these days (although it is admittedly a fairly different format).

    3. Nerds and engineer types aren’t always known for their tactfulness. Most of us have technical preferences based on a lot of deep rooted requirements or preferences and hardly ever just chose something for being pretty and easy. That can apply to the software we like, our design process, or our approach to hardware design/engineering.

      The internet already makes it simple for ordinary people to be jerks, so its no wonder a place like this produces such comments so frequently. It takes an active effort on all our parts to take a step back and attempt to be more kind in our comments or at least have the wherewithal to not bother posting a purely negative comment out of respect for others.

      Sadly many people just do not care how they affect others. Some people even pride themselves in how little they care about how their words impact people.

  5. Huh. Two “attacks” about using the RK3588. It is a competent 64 bit chip that can use 32GB of RAM and run 2 or 3 displays. People run ARM optimized Debian and Ubuntu and all the usual stuff with VM’s, NAS tools and OpenWRT. I use them every day and they make a fine desktop – that also has all the peripheral connections with I2C and SPI and a ton of other stuff. External interrupts timers and counters. Pretty darn flexible.

  6. Showing these kind of chops he could probably tackle a motherboard swap from one of these mini-PCs, N95, N100 or even 7840u etc 😆. Assuming they are thin enough, remove some ports etc.

    1. It depends. A conference room filled with techbros? Sure, there will be a lot of hype, lots of mutual admiration and nothing will get done.

      A conference room filled with embedded engineers working in, say, automotive? More likely to simply get a pat on the head and be told to get in line or get the fucׁk out because in the end it’s just a glorified dev board.

        1. Heck I’d hope your co-workers would be had he been the employed embedded engineer – its an impressive feat still, demonstrating a wider range of skills than absolutely required by your job. Which is very useful and interesting to you all. Nice to have the guy that knows enough about almost anything that you can find and work around the problems left to you from other teams and understand what feedback to provide for the next iteration!

    1. Yes, indeed. How dare someone present their serious work in an entertaining way that will get attention and highlight their efforts to a far broader audience??

      The abject horror.

    2. Are you sure you watched the same video as linked above? The video was overall extremely serious and to the point, and much more detailed than I expected coming from such a young developer. A little levity here and there is a good thing, it keeps the video from being stuffy and boring. Lighten up, Francis.

  7. Why the hell is everyone still building with the RK3588 like it’s some great new thing the damn thing is 5 years old and Rockchip still hasn’t released an updated version. You would think by now they would have released something almost equivalent of the Snapdragon x elite!

    1. You realise you answered your own question. “Rockchip still hasn’t released an updated version” rather says it all, there are not many vendors you can use as a small company/individiual and this one is still one of the flagships of those available options. So even though it is older than you might prefer…

  8. Thank you Byran for sharing your project, with source and all!
    You are extremely talented. Your skill set exceed most people, it exceeds mine by orders of magnitude. Please keep picking up new skills, do the project you enjoy to do. I would love to see more work of yours.

    To all the critics: shame on you. What is your excuse for NOT taking on such a project as complex and diverse as Anyon_e? If you don’t like the SoC, why not contribute, modify and add to the Anyon_e project?

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.