Arduino teams with Intel to produce Galileo

Posted on Sunday, October 6th, 2013 in Arduino, dev boards, News by the machinegeek

Massimo Banzi has announced the latest release in the Arduino line: the Galileo.

Galileo is a microcontroller board based on the Intel Quark SoC X1000 Application Processor, a 32-bit Intel Pentium-class system on a chip (datasheet — 920 page PDF.) It’s the first board based on Intel® architecture designed to be hardware and software pin-compatible with Arduino shields designed for the Uno R3. Digital pins 0 to 13 (and the adjacent AREF and GND pins), Analog inputs 0 to 5, the power header, ICSP header, and the UART port pins (0 and 1), are all in the same locations as on the Arduino Uno R3. This is also known as the Arduino 1.0 pinout.

Galileo is designed to support shields that operate at either 3.3V or 5V. The core operating voltage of Galileo is 3.3V. However, a jumper on the board enables voltage translation to 5V at the I/O pins. This provides support for 5V Uno shields and is the default behavior. By switching the jumper position, the voltage translation can be disabled to provide 3.3V operation at the I/O pins.

The Galileo is expected to be available in late November, 2013.

This entry was posted on Sunday, October 6th, 2013 at 1:07 am and is filed under Arduino, dev boards, News. You can follow any responses to this entry through the RSS 2.0 feed. You can skip to the end and leave a response. Pinging is currently not allowed.

26 Responses to “Arduino teams with Intel to produce Galileo”

  1. ken says:

    Any idea what the power levels of the board really are in idle vs. active modes? “1/10 of Atom” doesn’t help much :-P

  2. JBeale says:

    Interesting, that is a long way from the ATmega chip (…and also some distance in power consumption: “The recommended output rating of the power adapter is 5V at up to 3 Amp.”)

  3. GoogleMonster says:

    A look around indicates this board will be $200 each, not terribly enthusiast friendly. I dont think intel cares though. Now if they could put one of their Lenovo K900 processors on there.. then we’d be talking! (10% speed improvement & 30% power draw vs. present ARM architecture)

  4. Sleepwalker3 says:

    Good to see something coming out with selectable I/O Voltages, not nearly enough of that around. I am a bit puzzled about where Arduino are actually headed, what they’re aiming for, considering they start of with a pretty base level micro and have gone up to something like this. Now I wonder what OS it will run.

    Looks like they are sticking firmly to their mucked-up pinout. I thought by now they would have gone to a proper 0.1″pitch and just put out an adaptor for previous sheilds. There are bent headers which work the other way, so why not get the board correct and have headers or an adaptor shield if you want to run the old mucked-up pinout?

  5. Sleepwalker3 says:

    Hmm, read a bit more. I was thinking this was aiming to be something like a Pi, but apparently not. Seems like they are just talking about running ‘sketches’ (I really hate all these buzz words like Shields and Sketches and Platforms) like a normal Arduino. Haven’t seen anything to suggest it would run something like Linux, Android, Win or anything like that, but then I didn’t notice any video stuff.

    • JBeale says:

      They mention it has USB 2.0 host and PCI Express mini-card ports. So those will be supported from the Arduino environment? It will be interesting to see what that sketch looks like.

  6. Jack says:

    I think an important point about this is that it’s not an Arduino, it’s the first ‘Arduino Certified’ board. It’s being branded as an Intel Galileo, no Arduino logo. Which makes it an open question of just who is responsible for keeping the special ‘Intel’ targetting port of the IDE, platform libraries, and compilers in sync with official Arduino releases. Which could be a problem if whichever side is in charge loses interest, or isn’t good at keeping it up to date. I notice that the Galileo-supporting Arduino IDE already launched one version behind the regular one, and is hosted on behind click-through license agreements (even if following links from

    @Sleepwalker3: It’s an x86, with 256MB of DRAM, and a MicroSD slot. It’s going to be able to run Linux.

    But what’s kind of cool is that it only includes 8MB of SPI flash, so like you said it’s not targetting user-friendly OSes as its main use case, or not out-of-the-box. The board description on Arduino’s site (also hosted as a pdf ‘Datasheet’ by Intel) describes the flash as for the “firmware (or bootloader) and the latest sketch”.

    Sounds like Intel have written up a tiny ‘OS’ whose sole purpose is to handle uploads from the Arduino IDE, and run a compiled sketch, in whatever binary format the Galileo version of the Arduino compilers output. Actually, looking at the documents available, the getting started guide describes the ‘capsule’ firmware in the flash as a ‘Linux image’ at one point, so I guess it’s less an entire OS they wrote than one they customized, presumably with some sort of daemon that runs the sketches.

    They also offer a slightly larger linux image to use off of a Micro SD card, since they couldn’t fit the wifi driver into the on-flash version. So it can probably also boot into other OSes off the Micro SD as well. I don’t think they would have put a Mini-PCIe slot on the board if it could only run Arduino sketches.

  7. kerimil says:

    this and the arduino Tre puzzle me… if you want simplicity and ease of use there’s pcduino, if you want more I/Os and you don’t care about compatibility with your arduino ‘sketches’ then there’s the BBB. What’s interesting about it other than the fact that it’s x86 ?

    • Jeff says:

      Arduino seems to be clinging to the “simplicity” of the IDE, but that’s the weakest part of the whole system! I worry that the Arduino team does not have a clear vision of what “Arduino” is or wants to be, and risks being manipulated by large hardware vendors.

      A powerful IDE should be the top priority, but I haven’t heard the Arduino team talking about their next step in that arena.

  8. Rodger says:


    No video.

    It’s a lemon, same messed up and limited Arduino I/O. Look it’s time Banzai and company break away from that I/O config for the 32 bitters, I don’t care how much the base whines about it, it’s stupid.

    Mini PCI-Express? If you plan to use that for prototyping, you will pay through the nose.

    In addition the software can only access the on-chip SRAM. No ability to run any DOS apps or FreeDOS.

    Better off getting a x86 embedded board

    • hoeb says:

      Yeah but it can toggle its GPIO up to a whooping maximum of 500 time per second….

      • udif says:

        You wish. From what I’ve read, it’s more like 230Hz, not 500Hz :-(

      • Sleepwalker3 says:

        Hmm, sounds like the PicAxe 8 Mk I would be quicker for I/O!

      • hoeb says:

        The toggle speed is around 500 times a second which of course means the frequency of a wave is only going to be half that – ~250Hz.

        I did try to choose my words carefully so this would not be confused but seems just not carefully enough. WTF anyway. The point is that the GPIO is done through a I2C I/O port expander that runs @ 100kHz. What a joke…

  9. Sleepwalker3 says:

    @ Jack – Yes, it seems you are quite correct, it’s not actually an Arduino, it’s an Intel board that’s ‘Arduino Certified’ – whatever that means! Obviously they’ve made it aimed at the Arduino crowd who want to move up in the world, but to me it seems very disjointed, especially with no onboard video (that I am aware of anyway) – Mind you, having said that, I find the RasPi to be a bit disjointed too, I think that running such power without putting a lot of fast GP I/O out there is just silly. I know others will disagree and say you can add this and add that, but why? The boards are apparently aimed at people who want to make ‘stuff’ with them and yes, you can do a lot with them, but surely one of the key things, at least from my view, is to have tons of fast GP I/O to control stuff *without* having to add other boards, other processors, other devices which chew up I2C and SPI ports, etc. To me it seems like sticking a formula one engine in a low model 4 cylinder car with front wheel drive, razor blade tyres and standard brakes.

    It might be able to run Linux, but so can lots of things these days and you can plug a monitor into those other things if you want it. But there’s no Video (that I am aware of) on this and very limited *GP* I/O, both in terms of numbers and speed (apparently), so you’re back to adding this and adding that if you want that. It sounded great at first, but now I just see it as disjointed. Seems like a lot of computing power that, for the most part, can’t get out.

  10. Chuckz says:

    I hope everyone pushes for a “Mega” form factor shield / prototyping area on the board because since it is that much more powerful, it should reflect that in the number of pins available.

  11. DONALFONSO says:

    intel is a stupid dinosaur
    to turn on and off an LED I’ll still use the x86 instruction IN / OUT?
    and then the majority of pin bga (LOL O_O) are vcc and gnd.
    and stupid pin should be reserved for the various protocols. USB/ETHERNET/
    SPI/I2C etc. .. (not RAM because’ll explain further down) while I see that there
    are dedicated Cypress chip ecc..

    NO NO NO the whole shit

    now I give you an idea worth billions:
    for these toys disposable serving “Total integration”
    von numan said it would take Ram CPU and IO BUS
    and still holds the intel chip are different in 2013 ……

    ARM strength. Dinosaurs to death.

    sorry for the outburstby
    an Italian electronics hobbyist who discovered the internet and
    Ohm’s law. to 30 years and which uses Google Translate. to be understood on the Forums.

  12. sfsdf says:

    News about the Arduino Galileo but not about the Arduino TRE!!
    Hole in your news dangerousprototypes!!
    Contains powerful ARM processor! Lots of compute power. ArduinoBoardTre article on mentions 100 times more than the arduino Leonardo or Uno.

    Have a post about the arduino TRE!!!

  13. sfsdf says:
    It’s mentioned the arduino Tre can run ‘full Linux’, very interesting!!!

  14. sfsdf says:

    With being anble to run full Linux desktops it might be time to think about BIOS.

    Arduino with an Open BIOS would be awesome.

    Coreboot would be an enormous practical solution for this.
    Open BIOS, (with dualbios, necessary for tinkering).would make the platform very open.

  15. Tiersten says:

    You don’t need or want Coreboot as it is aimed at x86 and x64 architectures. A PC BIOS style firmware with an abstraction layer is not necessary for ARM anyway, as for an embedded platform you’ll be tailoring the kernel for that particular architecture and device. All you need is a bootloader and ARM already has two widely used bootloaders, uBoot and Redboot.

    The BeagleBoard uses uBoot so I expect the Arduino TRE to as well. The ability to run “full Linux” isn’t that unusual either since so does the BeagleBone, BeagleBoard, Raspberry Pi etc… The difference with the Arduino TRE is that there is an AVR bolted onto it with the forever misaligned shield connector.

  16. KH says:

    Datasheet says 400MHz. Minimum 100MHz. No information on power consumption that I can see. Perhaps done on a mature and cheap 65nm process. Hard to compare with any similar boards right now unless some benchmarks comes out. Yet another small board I can’t really repair; $60 too; heh, I think I’ll stick to <$1 MCUs for most of my projects…

  17. sfsdf says:

    And the joy of tailoring my custom kernels coupled with the fast technological progress of arm sounds a bit like being stuck in a loop.

    Did mean a wider form of startup solutions than traditional x86 bios solution.
    Too bad there isn’t standardization on a few things concerning arm:

    CPU’s do have at their core the concept of general purpose computing as an intrinsic feature! Don’t forget that.

    • Tiersten says:

      You’re missing the fact that the majority of these designs are for embedded usage where the ability to autoconfigure all peripherals and have a generic HAL is not actually wanted. Adding a bunch of logic to handle it just pushes up the price per SoC with no benefit for the intended purpose of that SoC. For most companies, the inability to easily swap out the firmware/software in their device is a feature!

      You’re after a more general purpose orientated version of the SoC which the Sitara isn’t. If you look at the TI Sitara page, you’ll see that it is intended for embedded devices only. For most ARM SoCs, you’re lucky if you can even get documentation without an NDA and from my experience, the documentation will have lots of things which aren’t documented publicly at all. We’re only just now getting good results from the reverse engineered ARM Mali GPU drivers despite the complete lack of help from ARM. The Broadcom Videocore is completely undocumented by Broadcom despite it being in the Raspberry Pi but some people have made good progress on that as well. Neither are actually ready for production code though. The GPUs are part of the secret sauce the distinguishes a particular SoC from another.

      Until you can get the manufacturers on board with good documentation and no need for mysterious binary blobs then your dreams of a completely open ARM platform with your choice of actual hardware is just a pipedream. It would be nice if it did happen but I’m not holding my breath for it happening any time soon. There has to be business case for this to happen and the benefits are pretty slim.

Leave a Reply

Notify me of followup comments via e-mail. You can also subscribe without commenting.

Recent Comments