Please create an account to participate in the Slashdot moderation system


Forgot your password?
Check out the new SourceForge HTML5 internet speed test! No Flash necessary and runs on all devices. ×
Music Open Source Hardware

Consumer Device With Open CPU Out of Beta Soon 99

lekernel writes "After years of passionate and engaging development, the video synthesizer from the Milkymist project is expected to go out of beta in August. Dubbed 'Milkymist One,' it features as central component a system-on-chip made exclusively of IP cores licensed under the open source principles, and is aimed at use by a general audience of video performance artists, clubs and musicians. It is one of the first consumer electronics products putting forward open source semiconductor IP, open PCB design and open source software at the same time. The full source code is available for download from Github, and a few hardware kits are available from specialized electronics distributors."
This discussion has been archived. No new comments can be posted.

Consumer Device With Open CPU Out of Beta Soon

Comments Filter:
  • by Hatta ( 162192 )

    Now I can patch my CPU. Oh...

  • It's only available from specialized electronic distributors never mainstream distributors. Why? Because mainstream distributors only buy from mainstream suppliers unfortunately.
    • I don't see anything about EMC or safety testing from an accredited lab, so I doubt this can be sold as anything other than a "development kit". That's not what you would find on Amazon or at Best Buy. If this thing really interests you, then they hassle of getting one will be worth it. This is more interesting as a general FPGA tinker box than for the stated purpose.

    • by cdrguru ( 88047 )

      The thing has no shielding, so it is going to radiate all over the place. It can't be used in a home environment - your neighbors would complain. It probably would pass UL testing because it runs off a power cube, I would presume, so it is just low voltage inside.

      Without FCC certification it is an "experimental" device and not going to be stocked by anyone as a "consumer" device.

      • by wspraul ( 594789 ) on Tuesday May 10, 2011 @09:12PM (#36089828)
        All wrong :-) It has no metal shielding because it is so well designed. We absolutely went to an EMI test lab to be able to classify it under CE and FCC regulation. Under FCC regulation, Milkymist One is a non-intential radiator and thus does not require an FCC ID. It is enough that the manufacturer verifies that it is in fact meeting the requirements of a non-intentional radiator. The entire test lab report (31 pages) is online
  • ok (Score:4, Interesting)

    by Osgeld ( 1900440 ) on Tuesday May 10, 2011 @06:31PM (#36088828)

    yea ok I will admit this is the first time I have seen an open source CPU, but that is becuase the rest of us would have grabbed a fpga and not wasted a bunch of time.

    I will also admit that this is cool as shit after calling it a waste of time, its a bit of both I guess

    • by Osgeld ( 1900440 )

      oh wait IT IS A FPGA, big FUCKING DEAL

      • What you got against FPGA bitch ? Chris
        • Power consumption, among other things.
          • Really considered about power consumption?
            You are completely Trolling.
            Even the new Spartan processors have power consumption options.
            Your common everyday x86 is a real HOG.
          • Re: (Score:3, Interesting)

            by wspraul ( 594789 )
            The entire video synthesizer runs on less than 5W. We didn't pick an fpga because we wanted to make an fpga computer. We picked an fpga because it allows us to make a spectacularly well performing and low power (!) video synthesizer. It can easily beat a multi-GHz Intel system.
            • by tempmpi ( 233132 )

              I doubt that, it might be faster than the multi-ghz Intel system without using the GPU, but using a GPU, even a onboard on, you should be way faster.
              You could build a fast DES cracker using that FPGA, but your task seems to be mostly regular CPU and GPU tasks. It is impressive to implement all this by yourself, but looking at the blockdiagram I would say a e.g.: Freescale i.MX51 or some Ti OMAPs with a small mcu for IR,MIDI and DMX512 could do everything as well or better than that SoC, including running at

              • Freescale iMX51 and TI OMAPs are completely proprietary, AFAIR. If at all you'll only get closed-source drivers for their built-in GPUs. That doesn't make them very sexy for such open-source hardware projects. Also I guess you'll soon run into real-time execution problems, if the GPU drivers aren't 100% perfect. Even with my ATI card I have these problems from time to time (something flushing GPU pipelines? no clue.). This FPGA CPU with RTOS kernel and custom-made 2-D acceleration will allow you to get
        • by Osgeld ( 1900440 )

          heh nothing, but shit man this is hardly the first FPGA open source computer, its not even really that practical, name 1 situation where a low powered SBC is really going to use TV input AND midi while driving a vga display, besides does anyone even use midi any more? We dropped it off our product line ~4 years ago and no one has noticed

          heck I just got a bag full of free standing midi voice boxes out of our dumpster for scrap parts today as they were cleaning out the storage room

  • What makes this great (and unusual) is that there's an entire SoC in that FPGA, not just the CPU core. And it's all Open, including the graphics.
    • Unusual? Really?

      Hasn't been around for 11 years already? Arn't most FPGA projects involving a CPU core considered a SoC? I mean, like most people won't have one FPGA be the CPU and the other FPGA be the peripherials, yeah?

      Heck, my last project in college would be considered a SoC on FPGA and that was like in 2003. (We implemented our own core, cache, and memory controller from scratch. Would have done a basic VGA output but ran out of time and couldn't afford to get myself a Virtex board after

      • by JackDW ( 904211 ) on Tuesday May 10, 2011 @08:03PM (#36089468) Homepage

        Indeed, hardly unusual. At the very beginning of, which was certainly around a decade ago, there was a project of this sort. "ORsoc" ran Linux. The CPU was an Opencores design named OR1200, with a completely custom instruction set and a fork of GCC/glibc to support it. Everything was open source: the peripherals, the CPU, the video drivers, even the USB and Ethernet cores.

        That SoC worked on FPGAs, but there were also ASICs, and I think it even turned up in some commercial products.

        I suspect that this project is probably reusing quite a few components from Opencores. That Wishbone bus looks awfully familiar...

        • You are right, Really Very few components from Opencores, actually Wishbone core is a modified version from the conbus core from opencores. But some cores developed by the main developer during the process, like the navre cpu and hpdmc controller we're published in opencores as well. Feel free to ask more questions here or at the IRC Channel at Freenode #milkymist also follow the development of the project by github A Quick Webchat link for IRC Chat here too: []
        • I do not make any special effort to sound negative here, but honestly that SoC did not really work on anything. Most of the stuff posted to Opencores is in fact half-finished, buggy projects, and Milkymist SoC does not use any other Opencores stuff than Wishbone (and still, this was because of LM32) for this very reason. Even the OpenRISC GCC/glibc toolchain was crippled with various major problems until recently. The OpenRISC RTL still is, but I can see things moving in the right direction. Maybe Milkymist
  • Meh. (Score:2, Insightful)

    by jmorris42 ( 1458 ) *

    Just slapping "OPEN!" on something doesn't make me wanna buy. For $500 I get a device that (someday when the software gets written) will process standard def video? And output VGA? Really? In 2011?

    Yes it is nice that everything is implemented in a FPGA and totally open. Perhaps someone will run with it and use these building blocks to make something interesting. But as long as an FPGA is the target it will never compete. Compare and contrast these features with what $25 will get you in an ARM. You ca

    • You apparently have no idea what this device is for. You should try taking a look at it before passing judgement.

      • by jmorris42 ( 1458 ) *

        > You apparently have no idea what this device is for.

        Yea, I read the site. But I'm still having trouble figuring out what it might do that a netbook with a hundred bux of USB devices stuffed into its ports can't. MIDI isn't expensive, NTSC video in isn't expensive. And the netbook will have more grunt than you are getting from the mmuless cpu simulated in that FPGA. So unless the idea (not mentioned on the site) is to have enough spare gates that video effects can be offloaded to special custom circ

        • To answer your question about spare gates, we are using about 44% of the FPGA resources at the moment. I would also question your remark about the compared "grunt" of a netbook, as many non-tech people I have shown the device to have spontaneously praised it for its reactivity and fluidity. Finally, some people are working on a MMU and even though it is of little use for my intended video synthesis application, you are most welcome to join them.
    • Price mainly has to do with volume. Also, VGA is still widely used today, and does not mean low resolution as the Milkymist One can do 1280x1024. We are planning to add a connector to drive HDMI displays at some point, which consists merely in wiring it directly to the FPGA as the Spartan-6 we use has the TMDS stuff built in, but unless we have the time and the development resources to get it done fast in the FPGA design, it is not a priority. (I work on Milkymist One)
      • Price mainly has to do with volume. Also, VGA is still widely used today, and does not mean low resolution as the Milkymist One can do 1280x1024.

        VGA is really quite good. The GMA950 will happily dump 2560x1920 over a VGA cable. Though, the graphics performance takes a big step down once the framebuffer goes over a threshold for reasons that I do not presently recall.

        There are not many minotors which are that big.

        Though presumably one needs a decent DAC to go that fast...

        Does the FPGA happen to have a DAC o

  • Professional DMX connectors have 5 pin connectors, not 3 pin.

    • Feel free to etch your own circuit board! It's open source!
    • by guruevi ( 827432 )

      Not necessarily, depending on the age of your components this can actually be a liability since the 2 extra pins were never standardized. Some (bad) designers have used it historically to carry a destructive current over it.

      • Yes and being a lighting professional, you probably already know that 3 pin 5 pin adapters are not hard to come by, don't you?
  • by mark_elf ( 2009518 ) on Tuesday May 10, 2011 @07:18PM (#36089142)
    Judging from what the screenshots look like, the "video art" it produces looks a lot like Winamp circa 1998. I'm part of the target audience for this thing and it looks pretty useless for making video. But if you shine some lights on it, it looks kinda cool I guess. If you tell people about the open source CPU it gets even cooler.
    • The material we posted online so far is not so great, but from feedback on real uses of the product, it does much better than what you describe. But it can be a matter of taste, too.
  • Open source? (Score:5, Interesting)

    by pem ( 1013437 ) on Tuesday May 10, 2011 @08:43PM (#36089690)
    I'm not particularly happy about my chances of legally
    reusing code that starts like this:

    //                           COPYRIGHT NOTICE
    // Copyright 2006 (c) Lattice Semiconductor Corporation
    // This confidential and proprietary software may be used only as authorised by
    // a licensing agreement from Lattice Semiconductor Corporation.
    // The entire notice above must be reproduced on all authorized copies and
    // copies may only be made to the extent permitted by a licensing agreement from
    // Lattice Semiconductor Corporation.
    // Lattice Semiconductor Corporation        TEL : 1-800-Lattice (USA and Canada)
    // 5555 NE Moore Court                            408-826-6000 (other locations)
    // Hillsboro, OR 97124                     web  :
    // U.S.A                                   email:
    • Re:Open source? (Score:5, Interesting)

      by pem ( 1013437 ) on Tuesday May 10, 2011 @08:52PM (#36089746)
      BTW, that license was from the tarball at []

      Before bothering with that, I actually tried figuring out the license by looking at Lattice, but other than reassuring verbiage about free, I came up blank when looking for an actual license: []

      And, of course, most of the Lattice junk in the source tarball, and the documentation at the milkymist site, can't even be retrieved from Lattice itself without registering and executing some sort of license agreement: []

      Lame. BTW, the main article links to [].

      Which links to the SOC code page [].

      • Please read the license from the Milkymist github repository: [] - "11. OPEN SOURCE SOFTWARE...."
        • It's mostly gobbledy-gook, that you would need to study way too carefully before playing with the code.

          In any case, unless there's something I'm completely missing, it looks like the milkymist guys were not supposed to share the code that I pulled that header from:

          Lattice hereby grants to Licensee a non-exclusive, nontransferable license to use the Software for Licensee's internal purposes only on any computer possessed by Licensee on which the Software is designed to operate, such use to be in accordance

          • by wspraul ( 594789 )
            We've done that like many others ("study the license carefully"), but you don't trust our results. That's your problem. The entire LATTICE.LICENSE applies to those files, not just section 11. Get your mind off of section 11, start reading at 1. Then 2. Then 3. and so on. If you don't want to do that, fine. Still your problem. Until you found out about Mico32 for the first time (today), the entire world knows since 2006 that Lattice has open-sourced the Mico32 core. Unfortunately they didn't pick a standard
            • First of all, I started out reading the whole license. Then I was pointed by one of your devs to section 11.

              Now you keep saying rude shit like "That's your problem." "Get your mind off section 11." "Still your problem."

              In short, you have been quite rude and snide on multiple occasions. AND YOU STILL HAVEN'T EXPLAINED WHY IT'S APPENDIX C OF THE LICENSE THAT APPLIES, AND NOT THE NON-OPEN-SOURCE PART.

              But let's get past that.

              The only "open source" parts of the license are described in Appendixes A-C.

              • by wspraul ( 594789 )
                Is everything alright with you? Hopefully Lattice can help you, I certainly cannot, sorry about that...
        • Re:Open source? (Score:5, Interesting)

          by pem ( 1013437 ) on Tuesday May 10, 2011 @10:23PM (#36090120)
          Forgot to mention that I did read section 11 that you point out to. It doesn't include all the source. There are plenty of files that self-identify as being under one of the licenses referenced by section 11, but the core CPU RTL files don't seem to fall into that category.

          Color me naive/fearful/stupid/untrusting/whatever, but when I see a license that covers both open source components and non-open source components, and a source file with a copyright notice that doesn't say anything about the code inside being under any kind of open source license, and in fact starts off by saying "This confidential and proprietary software ...", why the hell would I assume that the "open-source" parts of the license apply to that particular source file?

          • by wspraul ( 594789 )
            Because Lattice said so. Just keep reading a bit more, think, and keep posting. Eventually you'll get there :-) Wikipedia must have gotten it wrong too then []
            • by pem ( 1013437 )
              Anybody can edit wikipedia.

              I'm not going to spend any more time on that license. If you think it says what you say it says, spell it out for me, by chapter and verse. You can't just say that section 11 applies without saying why. Or alternatively, point me to a conversation with Lattice where they said it's all OK.

              • by wspraul ( 594789 )
                Sure, anybody can edit wikipedia, but you are not anybody. I have never said that section 11 applies, you speak to yourself and don't read documents from beginning to end. Not a good start for someone studying a license imho. I don't think section 11 applies, I think it's Appendix C, which you seem to not have gotten to yet. I think the starting point is here, "LatticeMico32 Open, Free 32-Bit Soft Processor" []
            • So I just asked Lattice's licensing department.

              Of course, them being lawyers, this discussion will probably be closed by the time they respond, but if not, I'll post the response here.


              Dear sir or madam:

              It has recently come to my attention that a public source code repository contains LatticeMico32 processor RTL files that have a Lattice copyright notice that claims the files are "confidential and proprietary software". For example, see:


              • by wspraul ( 594789 )
                Thanks a lot for contacting Lattice, I hope they can help you. What you ask them is hastily written nonsense, but maybe they can calm you down somewhat :-) I really think you should slow down a bit, think more, act less. But that's up to you of course...

                Who claimed that "section 11 of this license applies to those files"? Kristian Paul merely pointed to section 11. The entire license agreement applies to those files, specifically 1. 2. a. 2. b. 2. c. and so on. Maybe you read the entire license first and
                • MOD PARENT UP
                • by pem ( 1013437 )

                  Who claimed that "section 11 of this license applies to those files"? Kristian Paul merely pointed to section 11.

                  Are you listening to yourself? You act like I can't read and have to be spoon-fed (which may well be true), but that's completely crazy. Why the hell would somebody explicitly point out section 11 if it didn't apply?

                  Let's see what Lattice says, you certainly wrote to them in the most disturbing style,

                  What's disturbing about it? It's purely factual, and asks a question I would like to kn

              • by wspraul ( 594789 )
                The genesis of the source files seems to go through the Mico System Builder, then mico32_72_linux.tar. I was wrong pointing you to sections 1-xx, it should be covered by Appendix C. You said you had to register to download the mico32 stuff, but I am just downloading it without any registration. I will dig up more for you :-)
      • by wspraul ( 594789 )
        Are you sure the documents and sources cannot be retrieved without registration? Did you try to click on the links? It downloads for me (and I'm not registered)... []
        Can you try this one for example (random pick) []
    • No, that is exactly the definition of Open Source, and why the FSF shuns the term... It's no good being Open if it's not also Free (as in Freedom).

      I'll look elsewhere for the next hardware project to donate my time to.

      • by wspraul ( 594789 )
        Then you will like to hear that Milkymist licenses every new line of code under the GPL, version 3. That's our default license... If you want to help replacing the Apache-style licensed Mico32 core with a copyleft (GPL) one, that would be great. As of right now, 75% of the HDL sources of the Milkymist SoC are already GPL licensed, the other 25% are LatticeMico32 open source...
      • by pem ( 1013437 )
        There are legitimate things that gpl lovers and bsd lovers can agree to disagree about, but unless you can legitimately explain how a file can both meet the open source definition and be marked "confidential" please STFU.
        • by wspraul ( 594789 )
          Because LICENSE.LATTICE Appendix C supersedes it (and the 'confidential' file itself points to that). Why do you deliberately ignore all evidence that points to an official and proper open-sourcing by Lattice in 2006, most importantly what they say and offer for download on their website? It seems you have your mind made up, that's fine. We will continue to use the open-source Mico32 core until Lattice makes it clear that we (and the rest of the world minus you) misunderstood them.
          • by pem ( 1013437 )
            I wasn't talking to you. Rather, I was talking to somebody who thinks that "open source" is bad / GPL is good, and was apparently keying off the file header that I printed.

            To the extent that the file header is correct, it doesn't match open source principles, and what he is saying is bullshit.

            To the extent that the header is incorrect -- well, I'll reserve judgment on that. BTW, it is you who have been strongly asserting you know all about the license. I can't be "wrong" because I haven't figured it

  • by femto ( 459605 ) on Tuesday May 10, 2011 @08:46PM (#36089704) Homepage

    is the highly proprietary FPGA technology used to implement the CPU. FPGA partition, place and route (ppr) is some of most proprietary software on the planet, slathered in trade secrets and patents. The chips themselves are worse. Think of them as a type of processor (after all an FPGA is just a bit cruncher) with a secret instruction set and compiler (ppr). Xlinix (major FPGA company) want potential customers to sign an NDA simply to have their salespeople say more than "we sell FPGAs".

    If the Free Software community is to use FPGA's, as more than just a curiosity, first task is to design/build its own silicon and write its own toolchain. Then they come up against the proprietary nature of semiconductor manufacturing.

    I'm not belittling the Milkymist project, as what I describe above is a separate project. It's a huge project, essentially a reimplementation of 50 years of semiconductor progress, ultimately linked to the (seminal) desktop manufacturing projects that some have started. Imagine RepRap mk42 with semiconducting, conducting and insulating inks, printing circuits at the micro-scale.

    • Step by step !!,
      We’re aware of the current use of non-free synthesis tools. But it dint mean, we're happy with.
      So far the flashing process uses now free/open hardware and software, something that was not possible before.
      The history don’t end here, there are very smart people working on some replacement for this missing free parts, but is Work In Progress, and requires more people to join and develop around it.

      As in contrast the GNU compiler was no developed in a free/libre system from o
      • by femto ( 459605 )

        > Step by step !!

        No harm in dreaming! :-)

        Be careful to distinguish between synthesis and ppr. Synthesis is doable. PPR requires knowledge of the FPGA's structure as well as complete timing info. I agree that clues can be gleaned from the FPGA editor tools, but I don't think it's enough to write a PPR. (I could be wrong though, since I haven't tried it!)

        I'm keen to contribute, though I'm constrained by other things that take my time. One thing I do have is a complete MIMO capable reconfigurable radio []

    • by Arlet ( 29997 )

      Isn't it the same with all open source projects, running on a highly proprietary CPU ?

    • First, we are working on this, and your patches are welcome. [] [] FPGA companies are not as evil as you make them out to be. As a matter of fact, a large part of Xilinx's motivation about closing the bitstream is not to be evil, but to limit the damage that can be done from their (stupid and large) customers misusing the FPGAs. They still publish a lot and you might be surprised to learn, for example, that the ISE software has
  • The people who stand the most to gain from this are the same who gained the most from Linux --

    * IBM
    * Oracle (MySQL)
    * Every other large company that needed an easy way to implement a small product (e.g., LinkSys, a.k.a. Cisco)

    BTW, try getting any Linux open source code from IBM, you know, the stuff they are obligated to make public.
  • by smolix ( 133533 )

    I know this is going to be a flamebait. But before you flame me, consider the following: I'm researcher and get paid for what I do. I've released quite a few codes as open source and invented a bunch of algorithms which are not patented and used in many applications (think email spam filter, face recognition, etc.). And I've worked in industry and academia. For almost two decades. So I know both open and closed source.

    First off, ideas have value. As in Dollar value. Take NVIDIA for instance - they don't hav

    • First off, ideas have value. As in Dollar value.

      Not even draconian US "intellectual property" law supports that kind of idiocy.

  • But hasn't an open source consumer device already been released? []
  • by Animats ( 122034 ) on Tuesday May 10, 2011 @11:53PM (#36090482) Homepage

    Years of work, special purpose hardware, a price tag higher than an entire PC, and all it does is generate screen-saver like video wallpaper in sync with audio?

    If you're building technology for a rave, build something that makes the track spots follow the dancers. Something the dancers can play with. A Kinect might make that work.

    • by artor3 ( 1344997 )

      I think that if a Kinect saw a rave, it would have a seizure.

    • About the price tag: this has to do with volume alone. The only way we can bring it down is by selling lots of device. Traditional chicken and egg problem in electronics...
    • Ah and yes - the video input is pretty much to make stuff the dancers can play with. We have no good video footage or documentation of this however (which is part of the reasons while it's still "beta").
    • If you know how the unit/compositing setup works you can implement all kinds of additional software layers that will do just that. I have a Kinect interacting with Resolume 3 right next to me at the moment that I'm testing. My biggest issue at the moment is actually TUIOKinect being an asshole...everything else works great.
  • Try the UltraSPARC II instead, which is released under GPL, not some licence-proliferating legalese that's not actually open at all [].

    • We considered this option, but OpenSPARC is very resource hungry. It is a good design for a stand alone ASIC microprocessor, but in our case it is better to use a small and resource efficient CPU and leave the bulk of the calculations to dedicated accelerators.
      • Did you consider the LEONx cores? They are available as GPL code and are a great deal simpler and more FPGAable then the UltraSPARC II.

  • By the way, I will present the device tomorrow in Amsterdam: []
  • Interestingly, the OGD1 board produced by the Open Graphics Project is a significantly more powerful device. The problem is that very few OGD1 boards were produces, due to lack of funding.

    • All OGD does is a dumb VGA framebuffer due to lack of development/skills on the FPGA design. We went a lot further than that.

The world is moving so fast these days that the man who says it can't be done is generally interrupted by someone doing it. -- E. Hubbard