An Intel Atom CPU In The Raspi Form Factor

For years now, people have been trying to stuff an Intel processor on a credit card sized board. An x86 board that can fit in your pocket is an intriguing device – after all, that’s what Gumstix, the forerunner of the Raspberry Pi, were. Efforts to put x86 on a dev board have included the Minnowboard, the Intel Galileo and Edison, and even the Intel Compute Stick. These have not seen the uptake you would expect from a small x86-powered board, but that tide may soon turn. The UP board is exactly what you would expect from a Raspberry Pi-inspired board with a real Intel processor.

The feature set for the UP board is impressive for a credit card sized board; it’s powered by a quad-core Intel Atom x5-Z8300 CPU running at 1.84 GHz. The board comes equipped with 1GB of RAM, 16GB of eMMC Flash, Gigabit Ethernet, five USB 2.0 ports (one on a pin header) and one USB 3.0 port. Up also includes a real-time clock, HDMI, the same 40-pin GPIO pin connector found in the Raspberry Pi Model B Plus, and DSI and CSI connectors for the Raspberry Pi camera and touch screen.

To be fair to all the previous attempts at making a board built around an x86 chip that borrows heavily from the Raspberry Pi, there haven’t been many chips out there that have been suitable for credit card-sized applications. Only in the last year or so has Intel released chips suitable for an x86 single board computer, and the growing market of Windows 10 tablets bears this out. While it remains to be seen if the UP board will be a success, more than a few people will pick one of these up for a miniature Skype box.

133 thoughts on “An Intel Atom CPU In The Raspi Form Factor

        1. Yes, you’ll be able to buy it when you (& everyone else) help us hit our minimum order quantities. Unfortunately, vendors won’t just give you volume pricing out of the kindness of their heart or because you have an idea/post something on the internet.

          But if you would rather just ‘wait and see what happens’ then feel free. Just don’t expect the product to ever actually ship…or to ship before the price gets doubled. On the other hand, if you want to help reward someone’s hard work and see another dream/innovation get realized (that they’ve spent 16 hours a day working on for the past 2 years), then we will warmly welcome you as a backer.

          Alternatively, we are also accepting $55k checks from rich uncles or anyone else interested in seeing this project succeed. In that case, how many would you like?

          1. The way capitalism rewards your hard work is that I give you money and I get a good. The good you are selling with that post is… your warm welcome.

            I’m not buying.

            Crowdfunding passes all risk to the buyers, not the people bringing goods to market. If I’m accepting risk, I want the corresponding rewards as well.

          2. I’m not sure I follow. What additional reward would you like for $55?

            I think the risk of crowdfunding (for the creators) is lost on most people. I can’t really think of a riskier proposition than coming up with an idea, publicly broadcasting it to the world, setting an all-or-nothing threshold for its success, and seeing if people buy into it. If you fail, everyone gets to sit back and say “I told you so” and you’re left drowning in a pool of public humiliation, any potential investor deletes you out of their rolodex, and you question yourself for the next (whatever period of time) if it was all worth it or what you could have done differently. If you succeed, you get to deliver a product then go through the arduous process of turning it into a “real” business.

            Do you see it differently?

        2. Hi Chad, Jools, Velli, and everyone – I apologize. I haven’t slept much recently…or in a long time. I haven’t had a decent meal in weeks. I haven’t seen a lot of my closest friends in months. I literally wake up and do nothing but work on this project all day every day 7 days a week. It’s not a hobby or a past time. It’s my life. Needless to say, it means a lot to me.

          I don’t expect everyone to agree with what we’re doing or find it interesting or really care about us or our project for that matter. I do occasionally let my frustration get the better of me when I hear people badmouthing or second-guessing our intentions, our ability, our dedication, and, in some cases, our integrity. And for that, I am sorry.

          We have put more of our time, bodies, and souls into this than most would consider reasonable or even humanly possible. We’ve been told – and continue to be told – it’s a bad idea and that we should give up and do something different/easier for more reasons and by more people than I can count.

          Ultimately, if we can’t get anyone else to buy into the *vision* of what we’re proposing to provide, there’s really no one to blame but ourselves. And if that’s that case, it’s just something we’ll have to live with.

          So again, my apologies for coming off as a prick. My emotions got the better of me.

          1. I’d say the big difference between the RPi, Arduino, etc and the Snickerdoodle is that the former were not crowdfunded. It looks like if I buy a Snickerdoodle, I’ll be joining a community of almost 600 people (some of whom have very different boards), and have no guarantee of buying another one in the future if I need to. So I’d essentially be on my own, which isn’t the appealing to me.

          2. Thanks, Greg – that’s a good point. I appreciate that it’s a gamble (on both sides) – of course, it’s a total chicken/egg problem. The primary reason we went the crowdfunding route (aside from the fact that we aren’t all Broadcom executives with our finger on the ‘make 100,000 units now’ trigger) was to address that very issue: build up as solid of a community as possible right out of the gate so we can help people build great stuff and start aggressively expanding the community on that initial momentum of early backers. Of course, we are starting from “zero” and it’s not an easy task. Ideally it’s a totally shared risk but if people don’t see it that way I certainly understand.

      1. Ryan, I am truly interested in backing your project. Having said this, I would like to discuss why Snickerdoodle is having trouble gaining traction.

        Why was RPi so quick to grow, when Snickerdoodle hasn’t been? Well, RPi was targeted at at a larger demographic, which required a lower pricing point, and thus fewer features. Also, RPi had near exclusivity, with minimal competitors.

        Regardless, I think the largest issue with Snickerdoodle gaining traction stems from a disconnect in the products marketing.

        You have made fantastic product, You have differentiated it from existing offerings; but maybe too much. Your product is easy enough for a new entrant to find utility, but your marketing is aimed at the finite number of sufficiently advanced users who are familiar enough to program an FPGA. Why? Most people buying the RPi are unlikely to know that FPGA is an acronym, much less what it means, or does.

        So why not make the implied comparison by calling your product the ApplePi and calling modules Pi Plates? Nothing is more American than apple pie and going big.

        Afterward, you might consider clearing up the specification presentation by placing a lot of the more detailed information on secondary pages and providing clear and concise language that leaves no gray areas. Further, using a non-svg image was a poor choice for displaying the offering matrix. Why not just use html bullets or a table; displays better on phones in most cases. In any event, when selling to the masses, less information equals less chance for confusion.

        Ultimately, I hope your endeavor is successful; but if not, rebrand and try again, because you have a great product.

        1. The Raspberry Pi took off on the coat tails of the OLPC project and by drawing favorable comparisons to the original BBC micros. It’s only fair for new competitors to the Raspberry Pi to try and build off their market/mind share.


          1. There you go. There is precedent in this market to support this behavior too: RPi, BPi, OPi…

            Pi is kinda like Beer. There are millions of kinds of beer and whether its a ale, lager, stout, or even flavored, its still beer. Everyone knows what a beer is and its purpose. But, changing the name Beer would require a herculean effort.

            Why not sell the masses on the fact that you are offering high-point beer(FPGA Pi)?

          2. Yeah, that’s kind of what we were banking on: there’s this community of X million “makers” itching for something that enables them to take their projects to the next level and unshackles them from the limitations of R Pi and Arduino. The timing seemed perfect. But maybe it wasn’t…and never will be.

        2. Thanks, Donkey. We pretty much immediately realized that “FPGA” was a four-letter word and all but eliminated it from the campaign. Unfortunately it’s still in the video, which in hindsight probably wasn’t a good call since 90% of people are likely to make their decision within that first 2:30.

          You may be right about the name, but “snickerdoodle” probably isn’t the thing holding us back… I mean as far as “FPGA boards” go, it doesn’t get much friendlier.

          I didn’t quite follow re: the SVGs? The product images are of course bitmapped – we of course used SVGs wherever we could. But sorry, maybe I’m missing what you’re saying there.

          I appreciate the kind words. Just tryin to fight the good fight, ya know? In reality, we might not be able to do as much “good” as we thought we could when we set out – that is: provide professional-caliber tools at maker-friendly prices and give the masses the tools to build great things that otherwise would never be built.

          As much as I’ve/we’ve resisted it, at this point it’s looking like it’s no longer a question of *if* we’re going to have to raise the price after the campaign but rather *by how much.* Which is really unfortunate – that enough people couldn’t understand/realize the value of what we’re providing…either as a result of us not conveying it properly or people not really caring about building anything meaningful. Or maybe both.

          1. I wholeheartedly agree that the device is not holding you back, as it appears to be the most consumer friendly FPGA board in the market.

            However, I do think the name is holding you back; because it isn’t helping you. By choosing to keep this title, you are forcing yourselves to build an entirely new, untested, untried brand; which is contributory to your inability to attract backers. This is why I suggested that you might leverage the well-known Pi moniker; because it represents instant name recognition.

            Imagine the millions of kids who will receive a similar device this year on a holiday. Many of these kids will know what the RPi is because they have seen or heard of it. But how many will know what a Snickerdoodle is? None. What’s worse is, whose going to buy it if they don’t know how to explain it to the kid? If only there were a way to convey that it is like an RPi, but BETTER… Oh wait, (RandomName)Pi. Now they know what it is, what it does, and probably where to get started.

            This is how you reach those “masses” you talked about above. If you don’t change the name, you are effectively cutting yourself off from “the masses”, because you are catering to the finite class of people that will read through and understand the information you have provided.

            Finally, re:re:svg, two of three devices I used to view your CS campaign distorted the BMPs to near illegibility. I know you can’t control how individual devices present images to the user, which is why I suggested tables, lists, and svg.

            Again, best of luck with the product. I couldn’t imagine how hard you have worked on it.

    1. Would Hackaday please start putting a [Kickstater Campaign] warning label on its posts as they do for PDF’s? At least PDF’s often convey useful information … my personal experience with Kickstarter and Indiegogo campaigns has been much less satisfying. ;-)

    2. I thought that Intel Edison was pretty unapealing due to it’s over-minaturized form factor so I guess I care. It’s impressive that they can fit the things in such a small package but when you have to spend the price of a Raspi just to get a breakout adapter AND you lose all your size benefit to the adapter AND you can’t even get away with just one universal adapter, you have to pick from a group each of which only break out a subset of the potential signals…

      that seemed pretty sucky to me!

      1. Yeah you can definitely get carried away with miniaturization. We went out of our way to *not* use the super dense connectors so you could take the $55 version and jack a housing/jumper right into it and get prototyping without being forced to buy a baseboard. Allows people who don’t mind the extra bulkiness of 0.1″ headers or USB/Ethernet connectors to plug into a larger baseboard while still providing a “brain” that you can wire up and stick in a robot/drone/etc. that’s super lightweight and compact. There do end up being ‘compromises’ in both directions but it was the only real way to get the best of both worlds…

  1. I really wish boards like this or even a little bigger existed with expantible memory. For instance I have an Odroid U3 as the host board on my 3D printer which allows it to run Pronterface quite well, and the arduino “Dev environment” for making quick tweaks to the RAMPS firmware but 1 to 2 GB of RAM doesn’t really cut it for OpenSCAD or Sli3er when fed models of even moderate complexity or size.
    With a 64 bit CPU you’d really hope to be able to give it 4 or 8GB of RAM but with no SODIMM slots and 1GB soldered down or package-on-package you sort of hit a wall capabilities-wise long before you run out of CPU horsepower…

          1. I’m really starting to wonder what drives the success of all these R Pi/Arduino clones/knockoffs this vs. things like this: (full disclosure: I’m one of the project creators)

            I am coming to the realization (maybe a little too late) that “makers” really aren’t interested in building anything beyond a LED-enabled garage door opener or an automated cat feeder. We architected and priced this thing to give people access to something they’ve never had before without needing to learn anything new (yes, you can 100% program the thing using Python/Java/C/C++) and to open up a tremendous world of possibilities and opportunities.

            The idea was that with built-in wireless and a ton of I/O, you can essentially build *anything you want*. But I get the overwhelming sense that nobody really cares about that. It seems that most people don’t want to challenge themselves to create something different and they’d rather just have a cheap “computer” they can fit in their pocket.

            Am I wrong? Or are we off base with what we’ve done? I honestly don’t mean this as a slight to anybody making or working with these types of products, I really am just curious to hear people’s thoughts.

          2. Old Atoms are nothing like current Atom processors…

            Older Atoms (260, 330) were quite pokey, current quad-core Atoms at nearly 2 GHz are nothing like them – much more useful.


          3. @Ryan – I think I’m interested in building all sorts of interesting stuff. My time though is very limited. Anyway… SBCs like the Raspi stimulate the computer geek in me far more than the maker geek. Right now my time and money are going to other projects but soon I intend to buy a bunch of SBCs, Raspi, BananaPi, Intel something or other, whichever one hits the performance I need at the lowest price at the time.

            I don’t really want them for their GPIOs and such. I just want to offload a bunch of server processes that currently run on my desktop 24×7 as well as fill a couple of other server goals I am not running yet while maybe saving some electricity. If I could get something for a RasPi price that traded all those GPIOs for real (non USB) ethernet jack and a SATA port I would jump on it!

            It’s not that I don’t care about all those cool Maker, IoT, etc… things one can do with them. It’s just that I would rather do that stuff with a traditional micro. The micro does one thing, real time. The SBC runs a multi-tasking operating system. The micro is therefore better at most of those maker applications anyway! I suppose I could install a RTOS on an SBC but why? There are all sorts of sollutions both wireless and wired for connecting a micro to a computer. I don’t need the power of a whole separate computer everywhere I might use a micro!

            Now the new thing seems to be SBCs with FPGAs. FPGAs are on my some-day list. Some day I would like to lern about them. Some day I probably will. But.. when that day comes, why do I need a new computer with the FPGA built in? Wouldn’t I be as well off with a cheap SBC development board that plugs into my real computer via USB? Then when I actually build something with it I can just get a new FPGA and plug it back in to that same computer to start the next project.

          4. Interesting points, me. Minus SATA (which we’re working on a solution for, along with USB 3.0), the architecture we’re using pretty much perfectly accommodates everything you’re looking for – a flexible “brain” that can be plugged into different baseboards to give you what you want. Is it the cheapest solution if all you want to do is plug in a keyboard and display. No. Is it the cheapest (and most convenient) solution if you want a keyboard/display/gigE in one application, a basic breakout board in another, and a gigabit networking router in another? Yes.

            There are free, lightweight Cortex-M uCs you can slap in the programmable logic already. You can run Linux, etc. on the uP. We’re starting with easy I/O reconfigurability as a way to pick off the “low hanging fruit,” as that is the first thing you’ll probably want if you’re building a mechatronic system, but if you don’t need the I/O, use the FPGA to accelerate the $55 web server you’re running in your bedroom. No reason to run an RTOS/bare metal if you don’t need to; it’s just one of those things that if you need it, you already know you have to have it…

            FPGAs have (deservedly) become known as a dirty “four-letter” word in the maker world. If we get enough people together to get things rolling, our goal is to change that and have the technology, architecture, knowledge, and approach to do it. And we just might have another trick up our sleeves that will blow the roof off this thing…

          1. It may depend on the details of the memory controller. Sure, on a PC like a typical consumer whore, you “just add memory” and it works. As long as, say, that memory has the SPD chip on it to tell the controller it’s configuration, and the controller can read and support it.

            How much do you really know about memory? Should you be talking?

          2. It may require a new part with new pin-out, that requires a new/different PCB.

            Probably need to exhaust inventory of 1 gig units to meet production quota and meet contract terms with factory building the boards.


          3. The SoC memory controller is designed for 2GB max memory.The memory pinout on the datasheet should show the extra address pin for next density memory from the vendor. As for the memory timing parameters,, they should be very well documented. The board designer(s) get to pick the part and not the end consumers, so SPD isn’t necessary. Chances are that they might need to tweak a few parameters for the memory organization and may be timing, but that should be done in a couple of weeks.

            I would also be very surprised if the reference designs don’t already recommended a 2GB part with the necessary schematic, SoC parameters and recommended memory part already documented.

    1. There are Min-ITX motherboard that are far more expandable at a far cheaper price. They are larger and doesn’t come with GPIO. There are also the PC104 form factor SBC for industrial applications, but those are expensive.

      1. Yeah, that ends up being upwards of $135cdn. I should have been clearer, there, as the general assumption is USD. My apologies.

        The Odroid XU4, with an octacore BIG.little w/ 4 2ghz A15 and 5 slower low power A7 cores, 2gb ram, gigabit ethernet, USB3, and supporting eMMC storage in addition to uhs micro sd cards, is $75USD, or pretty much exactly $100CDN. It’s pretty much objectively better, substantially cheaper, and has Android and Linux OS’s ready to go.

        Can you run windows on this (bonus points: would it run well enough not to be terrible?) I’ll fully admit, I’m not up on newer Atom performance, after some bad experiences with them in the earlier days.

        I don’t want to rain on their parade. More SBC’s are great. I’m just struggling to see what this provides that’s not already covered, better, by existing and cheaper devices?

          1. No real comparison to the odroid above, though. I’m not gaming with these things, so it’s not really something that’s meaningful to me in terms of performance. Being able to run mobile games well doesn’t mean much to me.

            Benchmarks show it’s on par with devices like a Note4/Nexus 6. That’s certainly decent, and put it into the same general area as the Exynos 5422 in the Odroid. The 1gb ram, however, is absolutely a severely limiting factor.

            Basically, you’re paying a 35% premium for x86. Maybe that’s worth it for some, but it’s definitely not a selling point for me.

            With that said, if you can simply – effortlessly – install Windows 10, and get good performance with it, that’d be a huge plus for many people. Personally, I’m very happy being able to run Android and Linux, but I could think of use cases for a full, regular Windows install on something like this.

          2. “Personally, I’m very happy being able to run Android and Linux, but I could think of use cases for a full, regular Windows install on something like this.”

            Android runs on x86 (among other architectures), doesn’t it?


          3. I didn’t mean to imply Android wouldn’t run on your board; just that you were adding Windows (Can I just install regular windows 10 on this? I’m still unsure) for $35 usd more – and less ram.

            Are you providing an Android build at launch and supporting that?

          4. Something got cross-wired, I am not involved in this project – I was commenting on the ability of Android to run on x86 architecture machines, I’ve seen OEMs supply tablets with x86 CPUs running Android.


          5. The videos for games in windows pretty much is the answer for “Can you run windows on this?” It also means there is some memory left for graphic textures etc (not a whole lot), but should be enough for non-demanding stuff.

          6. I’m probably not phrasing this correctly.

            Please understand, I’m not trying to put this device down; more, I’m trying to understand where the value is in terms of price vs. competition.

            That you can run Windows on a different device that uses the same SoC isn’t really relevant. Here’s where my phrasing is misleading.

            More accurately: If I take my Windows 10 install media (USB key), plug it into this, can I simply install Windows 10 on it? Or will it be more complicated than that? And from there, will this be – for all intents and purposes – a windows 10 desktop PC in terms of hardware/software compatibility? Or is it more “It’s possible to run Windows on this, but not the normal desktop build.” If it’s the later, is it available?

            I’m still not sure it’s worth the loss of a GB of ram and the extra $35, but it’s something if one can just install windows without any hassle. I still haven’t gotten a really clear answer on that, though – Windows being shipped stock on a tablet isn’t indicative of whether or not this will Just Work.

            I’m also willing to accept “This is a dev board, you’ll have to figure it out” as an answer, though that gets it put into “this product is not for me” box. I’m fine with that. I just wish this was a bit clearer :)

          7. The tablet being discussed run COTS Win8.1/10 without effort, and I believe it was also made to run Linux, as a testament to it’s ‘standard-ness’.

            That doesn’t answer the question about what this particular board can do, but it is a proof of concept that on standard hardware incl. this quad-core Atom, Win 10 can run AND provide an application workspace with 1 Gig RAM.

            Since Windows Vista (which still has several times the end-users desktop Linux does, btw) Windows has gotten better and better army running in reduced memory configurations.


          8. So, as this is your build: If I take my Win10 install media, plug it in, will it just install, and run, and be a tiny Windows 10 PC? Or is it not that simple?

          9. You’ll have to talk to the vendor of the board to make sure that it comes with windows 10 drivers for all its hardware. Windows 10 is still in beta. I read through the forums of one other tablet that didn’t have the blessing of windows 10 and some of people had to go back to windows 8 to get their peripherals to work.

            A non-shipping first version kickstarter project with no track record is considerably worse than an actual shipping and working tablet from a well known computer vendor.

  2. Why are people so eager to sponge on Raspberry Pi’s success?
    The Pi is a really nice and very cheap device, but the physical layout is horrible, albeit better in the B+ model.
    But still all connectors are thrown in with a shovel, so what is the use of the pcb being small, if the end product ends up with wires and connectors going in all directions looking like a giant spider – and the supply of decent boxes for the Pi is scarce.
    So please, make some small computers, but not so small that connectors cannot be aligned at one side only, so casing would be much easier.
    And an other important thing I am missing is the Sata connector.

      1. The original Banana Pi has a SATA connector, it’s why I ended up getting one. The original Pro version even has built in WiFi. Sadly they seem to have scrapped SATA in the Banana 2. There are a few others (with arguably better specs) with SATA but I really couldn’t find a decent source for any of them. Shortly after I ordered my BPP the price on them went down, go figure :P

    1. There are cheap injection molded cases for RPi at below $10 from China and they look quite nice. The RPi connectors are very sloppy as as they point at every direction. I have made a point of lining up my connectors for my HaD projects even when it means extra layout work that few would appreciate.

      There are min-ITX and other standard form factors if you want a standard PC for a media box/router etc. (see the x86/x64 boards linked at that page) (ARM based)

    2. The connectors on the Raspberry Pi weren’t optimised for integrating them into small devices, they were placed so that kids and n00bs could easily hook stuff to them. The fact that hackers and advanced users bought most of them was an indicator that there is a market for inexpensive yet capable and, most important – easy ramp-up and buckets of support.

      1. The most important feature of the Raspberry Pi is it’s price, quickly followed by it’s ability to run Linux and third by it’s size. I suspect most people never actually hook anything up to the header pins on the Pi because they are using it as a streaming device, web server, or really teeny-tiny desktop computer.

        I suspect a great number of Raspberry Pis are sitting idle in junk drawers, their owner losing interest in their impulse purchase.


        1. Both excellent points, Ken…and Ken? I get that exact impression: people generally don’t care about hooking stuff up to the (very few) I/O they have on there and the form factor seems to not really be a factor either. If those things were truly what mattered to the R Pi audience, we would have sold 100k of these in a week (which would also take care of the ‘lack of community’ problem) seeing as it’s the same price as an R Pi w/Wi-Fi & BT:

          1. This is true. I’d definitely argue that the majority of people who actually use Raspberry Pi’s are using them as media players, mini computers, an that sort of thing. The advantage of them having GPIO, however, is you create a bigger community: The more devout maker sorts can build interesting stuff using the GPIO, and at the same time a hell of a lot of other people are using them as servers and whatever else. Also, at $35, you’re moving into “Why the hell not give it a try; it’s basically free!” territory.

            Thus, you get a larger community, allowing you to hit that critical mass where the community takes over software development.

            A purely Maker board, like your snickerdoodle, is going to have a much, much more limited appeal.

          2. Indeed. I’ve heard that a lot of R Pis (from a volume perspective) are used in retail displays and other signage as well, which is interesting.

            Guess we just had higher hopes for the maker community; although I’m starting to believe (or I refuse NOT to believe) that people might be into building/messing with more advanced/cooler stuff if we just *showed* them what’s possible. Clearly people who’ve backed the project to date “get it” and know what they’re doing/realize the value – and, admittedly, we probably haven’t adequately conveyed that value to the greater audience.

            We’re really looking forward to getting these into people’s hands because 500, 10000, or 1M creative people playing around with this stuff are likely to come up with *way* better projects and things to do with snickerdoodle than we ever will…

          3. I’m sure some would. I don’t know how big that market is, really – I think if anything the Raspberry Pi draws people because they bought it for some simple task, then had a few kicking around, saw a project they could use what they already had to do something genuinely cool and tried it.

            I suspect you’re going to have a harder time selling the device as a more dedicated maker board, though.

            I mean, I’m a moderately hacky person – nowhere near what most people here do – and while I’ve got a mountain of SBC’s, they’re pretty much always bought for a “normal” purpose first, then repurposed into more interesting roles later as they get replaced. I kind of think I’m probably more representative of the bulk of Raspberry Pi folks, then.

            My first Pi’s where bought as an entry level file server and a pair of HTPC’s. The fileserver was replaced by a full PC (needed many SATA connectors, as it grew fast), the HTPC’s by Odroids which are vastly superior boards, and at the same price – if lacking the breadth of community the Pi’s have. Since, I’ve used the Pi’s for a huge variety of random other projects. In fact, I only ever used the GPIO’s on any of my SBC’s after they’ve been taken out of duty from more “traditional” roles. Those “traditional” roles are basically the hook, an easy reason to tell myself to buy one. Once I’ve got it, then it’s on hand for other projects.

            I mean, I look at your Snickerdoodle page, and think “This board has a hell of a lot of potential.” But I have no idea what I’d do with it. Not that I don’t think there’s something that I can do with it, but it’s a chicken and egg problem: If I come across a project that could use it, it’s also quite likely I could make do with a leftover Raspberry Pi or ODroid. The Snickerdoodle is simply better in most ways, but I don’t have one in a drawer.

            Anyways, none of this is meant as criticism; just my thoughts as a random person who buys a bunch of these sorts of boards.

          4. Honestly I think your biggest problem with the snickerdoodle (just in my personal opinion) is that no one knows about it. I mean outside of the maker/hacker world how much coverage has it really been given? Not much from what I have seen. And even then, how many maker sites have covered it and didn’t just mention it in some small link off to the side?

            Another major problem I see (and I mean no offense) is that you greatly overestimated the ‘maker’ community. I think the maker community is way overrated. I don’t see a lot of true makers. I see a lot of assemblers. Assembling premade parts that fit together like a nice little puzzle. And while there is absolutely nothing wrong with that (I think we are all guilty of that to some extent) it seems a lot of the things people are assembling is absolutely pointless and gimmicky junk that can be controlled with copy/pasted code on the Raspberry Pi. But I do think the true makers that would use the snickerdoodle are out there.

            With all that said, I have not bought the snickerdoodle despite hearing about it early on. There are a couple reasons. Firstly, I have been burned way too many times with crowfunding the last couple years. Not that I don’t trust you guys, but 100% of the projects I have invested in have failed or delivered well under what was promised. My track record is so bad I can’t justify it anymore. Secondly, while I think this board would be really fun to play around with, that’s likely as far as it would ever go for me. I have several projects in mind that would work great with this. But at this time I no longer have the time or the disposable income to make those projects a reality. I hope you guys find enough people that do and I wish you guys a success.

          5. Thanks, Casey. If only it were easier to get more exposure. Unfortunately more people seem to like reading/writing articles about Intel’s latest Arduino clone than about something new and different. We’ve been riding the press since day one – the few we’ve talked to have been awesome and they totally get it. Sadly, that is not the norm…

            Re: the maker community, I’m quickly (and painfully) starting to come to the same conclusion.

            And I totally feel you on the crowdfunding front – that’s a pretty rough track record you’ve got going there (please don’t back any projects I do ;). Your final point about time/money to actually make a real product really might summarize it best. While we might be making it more affordable/doable than ever before for the “average” person to build something out of this world, it’s still no joke and you have to be pretty dedicated to stick with a project like that to the end (as opposed to say a breadboarded garage door opener tied to an LED).

    3. Unless a competing board has a strong existing community or a major unique selling point, then they need to have a similar form-factor/layout to even get in the door. Having a similar layout allows them to utilize existing Pi accessories. Accessories that would otherwise never be created for their board.

      It is extremely hard for competition to even get noticed in this industry as the Raspberry Pi has a larger community than every other board combined, several times over. And neither the community nor the foundation are very welcoming toward competitors.

      1. But I thought they were a non-profit and only cared about education ;)

        Gotta envy having a community that will defend you to the death no matter what. Kind of like Scientology, speaking of non-profits…

        1. I don’t have a whole lot of nice things to say about the foundation or the community. And after being recently told by some community members that the BeagleBone Black was a knockoff of the Raspberry Pi… I don’t have a lot of faith in that changing. Considering how little the two boards have in common it is hard for me to even comprehend that logic. I’d ask how they came to that conclusion on the forum but I would probably get banned for negativity and mentioning a competing board by name.

    1. different architecture, arm vs x86 – but speed really isn’t the key selling point, In my opinion the real selling point is just it being an x86 processor… which means i can run windows, linux, bsd, (potentially) osx plus being able to leverage the massively larger base of programming languages and programmers that work on x86 platforms.

      1. Which programming languages would that be? C/C++, C#, Java, Javascript, Python, PHP and mostly all othere interesting programming languages out there are CPU agnostic, so what programming language would this be a door opener for?

      2. A modern/current quad-core Atom CPU running at 1.86 GHz is a very capable desktop replacement processor – suitable for running, for example, the current Office 2013 suite very reasonably… It’s not a gaming processor by any stretch, but a good, web browsing/email checking processor with general office apps working fine.

    2. The gigabit ethernet, USB3 and eMMC flash all make it quite compelling. You should also be able to install windows on it out of the box, which will appeal to more people – if they manage to get a education windows agreement for a desktop distribution that would really be a killer feature.

      1. Good to see that this particular part has an Intel GPU instead of the PowerVR GPU that some of the Atom parts used.
        Depending on how open the BIOS is (can you run Coreboot on the thing?) this could be a good thing for people who want a truly OPEN small-form-factor computer (all of the ARM competitors rely on closed source blobs for GPUs and things)

  3. and yes an x86 board would have been nice earlier, to have an embedded windows ce5, 6, win7 or even win10 may be useful, we use embedded windows modules that cost far more (150$) and are not opened source.

      1. Yeah. The Minnowboard is pretty underrated. Though admittedly I don’t have one myself because I am not convinced the companies backing it (intel/circuitco) are committed for the long term. The lack of a community forum and generally quiet community doesn’t help my concerns either.

  4. The form factor or pretty close to it has been around for decades. PC-104 form factor used heavily in industrial and scientific with gobs of IO and other fun.

    all of this is newbies on the block finally catching up, I even think I have a couple of 386 based PC-104 processor boards in the basement junk box.

      1. Apple/Oranges? A good share of those WinXP machines (long unsupported and malware infested) are laptops which might or might not be connected to external displays via VGA or DVI. In the previous decade VGA has only be used if DVI were not available (e.g. some older projectors) or if users are oblivious of the issue and use really low resolution displays. VGA is dead and buried. Don’t dig it out again.

        1. There are many, many people running systems that include neither DVI nor HDMI interfaces.

          How many of those older laptops rely on VGA connections to overhead projectors?

          How many people are still using CRTs at home on older hardware?

          How many current model tier-1 servers include a VGA console, not DVI or HDMI?

          How many kiosks run on older hardware with touchscreen CRT displays?

          VGA is not as rare as you seem to insist it is/should be. It serves a purpose and a VGA connection can support a high-red display, and can run great distances without expensive repeaters (or ‘signal boosters’).

          You understand VGA is both an interface standard AND a name for a given resolution (640×480) – I, of course am describing the interface standard that relies on analog signals over an HD-15 connector…


          1. It’s not rare, but it should be. It saddens me when I go to work every day and see laptops in docking stations that support VGA, HDMI, and DP attached to monitors which also support all three but connected with a VGA cable. It supports high resolutions but it makes a mess of the signal on the way.

            It does have some specific use cases such as for hardened applications (VGA will happily display in crappy RF environments where it’s digital brothers drop out unless you go for HUGE bulky cables), but for everything else it should be considered legacy.

            Unfortunately it isn’t.

        2. I think a major reason VGA still sees use is because a lot of people have extra VGA monitors lying around. Their HDMI monitors are already being used and although you can use a TV, it certainly isn’t convenient.

    1. Many times, yes. Maybe not the rich people , but this kind of board will be used with small builds ( 15pol LCD with HDMI ? have not seen one yet ) , places where it is uneconomical or unpractical to change displays ( factory equipment, etc, where the displays are of some special kind ) , etc.

      Around here, a lot of displays is VGA -only or VGA+DVI. Changing to a HDMI-only scenario would be costly and people do not want to pay a lot when their old displays are still very good.

      And I ask the tft display connector because it is much cheaper to get a “dumb” display, like replacement parts for gps units, then one with controller ( SD1963, etc ) .

    2. I run the IT Dept in my office… VGA is still VERY prevalent even on new machines because Dell, etc still provide a digital port like DP, HDMI or DVI and an Analog VGA port. We need to run dual monitors so nearly every other screen here is connected via VGA

  5. I have a tablet that uses the quad Z3xxx Atom processor. It has only 2GB of RAM and eMMC, but it is quick snappy for casual use under Windows 10 x86. Running x64 Windows would use up more RAM and probably make it run slower because of that.

    Its much larger aluminum back (they use the case as heat sink) can get warm just for usual web browsing/video/RDP. That tiny card can get pretty warm as it with the much smaller surface area. It’ll get even worse if you stuff it inside a box without forced air flow.

    1. Why people aren’t building wireless into every new maker board these days makes no sense to me. Those dongles are total crap and make it *really* convenient to actually use it in a system since they prevent you from hooking up an external antenna….

        1. For sure. We threw a chip antenna on there along with a u.FL so once you actually put it in something you can install an external antenna outside your enclosure. I don’t see that happening with a dongle.

        2. Typically these ‘maker boards’ are very space- and price-constrained – adding WiFi in the base design would increase the physical size and price of the board.

          Also, let’s not forget the FCC – wifi adapters need to be approved for their intended application, which further increases the design effort and cost involved in the development process.

          As a historical note, back when people hooked computers to home TVs very few vendors included TV Modulators (converting video to a TV channel), instead relying on third-party parts like Sup’R Mod.

          Also, TI got forced out of the market when their TI 99/4 computer included a component (power supply? TV Modulator?) that failed to get FCC certification – a warehouse of units sat idle for months, unable to be sold until the system passed FCC certification testing.

          Simply ‘adding wifi’ is more complicated than adding a chip to the board in most cases.


          1. Ha you nailed it, Ken. To counter my own original point “Why people aren’t building wireless into every new maker board these days makes no sense to me” – after going through the exercise of integrating Wi-Fi on board, I know EXACTLY why most people/companies don’t do it: it’s hard and it’s expensive. Selecting the right part and carefully tying it into the design was miserable (and increases the cost of the PCB fab as you’re now introducing impedance-controlled layers, etc.). Not to mention Bluetooth, which requires a pre-certified, closed software stack… Yeah, we’re probably not charging enough haha

          2. …and woe unto the ‘maker board’ that picks the wrong wifi chip, the wrong choice can severely limit OS choices for the board and render the board useless if support for the chip goes away.


  6. I suppose if your mission is to build a really small Windows PC this might be interesting. For me it is a dud. I would much rather have ARM architecture and my interest is in embedded projects at a low price point. This is about double the money of a BBB (my favorite) or the RasPi — which gives this a complete pass. Nothing magic about x86 for me.

  7. for years now?
    for years now hobbyists have been ignoring PC104 and ComExpress both of which are available with I7’s,
    So congratulations once more for a poorly written ill informed shill promo for some new crappy attempt at filling a hole overflowing with better options

  8. Get rid of the DSI and the Ethernet port. Add WIFI. I couldn’t care less about the CSI port, but the kids will get all oissed if you take it out. It needs at least 2GBs.Don’t bother with uSD, go for either SSD drives or at the very least Compact Flash for better speed than the SSD. But no really, SSD port please.

  9. So, is the target market the maker community (hackers) or people wanting to throw together a small cheap PC or media player? I looked at the Pi and chose the BBB instead because it has more IO pins. For me it is all about IO pins. The new Pi versions are more attractive with more IO pins. I am looking for the cheapest little SBC with ethernet and lots of IO. I could care less about Sata or Hdmi — I’m not trying to build a linux computer and I certainly don’t want to run Windows. But that is just me, maybe there is a big market for itty bitty boards that run windows. The main thing is that the guys pitching this need to be clear about their target. If I didn’t want ethernet, I would go with some ARM based arduino like thingie.

    1. Tom, I don’t mean to ‘spam’ this thread as I’ve probably already left too many messages, but if it’s and embedded board with tons of I/O you’re looking for, you should take a look at this:
      Sure there’s no Ethernet, but there *is* Wi-Fi. And you can easily add (gigabit) Ethernet either via piSmasher or via your own baseboard. BTW, you can check out our GitHub for a reference design on how to do this yourself:

      1. Pretty funny. I read this less than a minute after doing a google search “snickerdoodle fpga” and skimming over the link to the crowdsupply page. First impressions are. Dude, look at all those IO pins. yeah, baby! And then there are worries about the FPGA learning curve (which I have been told I should not be afraid of, but this is my world, I have done PALs, so how hard could it be (haha)). But then I say .. wireless – no ethernet … bummer, I dunno. Well, just first impressions.

  10. And to everyone here – esp Casey, Ken, Donkey, and Derrick – I just want to say: thank you. Your comments and feedback have been very helpful and I appreciate the time you took out of your day to respond honesty (and sensibility). Guess one might call this whole process “trial by fire” and we’ll see where it all leads but I truly am thankful for everyone’s input and words of encouragement.


  11. Hi Everyone,

    In the same era, hundreds of people have similar ideas in the global market. But every team creates different things.

    We’re the Latte team, and we create Lattepanda. Lattepanda is the first computer board pre-installed full edition of Windows 10 with $69 price!

    Lattepanda is featured with Quad Core 1.84Ghz CPU, 2/4G RAM, 32/64GB eMMC with WIFI, Bluetooth 4.0 and USB 3.0.To make it suitable for a wide range of application, it’s designed with onboard Arduino compatible processor.

    We’ve finished some application with our sample boards! And have some great feedback from our partners for testing. If you’re interested in our product, please check our kickstarter preview page! We’re going to launch our project next week.

    Preview page:

    Hope you will support our project and any suggestion of the preview page will be appreciated! Thanks.

  12. Just found out about this thing today. Pretty hyped. Seems to be perfect for a project I’m starting with the Intel RealSense camera, so the USB 3.0 and the intel quad core chip are nails in the coffin for me. I mean how long has USB 3.0 been out 6.. 8 years? yet there are still only a handful of SBC’s that support it. I was beginning to think I would have to sell my soul to the devil for a SBC like this! :):))))

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.