Catch up on stories from the past week (and beyond) at the Slashdot story archive

 


Forgot your password?
Close
typodupeerror
×
Technology

Free 32-bit Processor Core 156

Posted by CmdrTaco from the but-who's-gonna-stamp-it-out dept.
Harry Houdini writes "EETimes has a story about how OpenCores is offering a free 32-bit processor IP core in a move that could undermine such commercial IP licensors as ARM and MIPS. "
This discussion has been archived. No new comments can be posted.

Free 32-bit Processor Core More

Free 32-bit Processor Core

Comments Filter:

  • VA Linux's commitment to the community (Score:1)

    by Anonymous Coward

    This is great news, but I don't see how it could be useful to most of us who don't have chip making foundries in our homes. Fortunately, I have a solution.

    With the recent merger of VA Linux and Andover.Net, Slashdot, and by extension, the Slashdot community, finds itself under a new company, VA Linux. Now VA is a commercial organization, and is in the business of making money. Fine, but here in the Free Software world, that's not enough. You also have to pay your dues, so to speak, to the community which you make your money from. And this is where my plan comes together.

    VA Linux is primarily a hardware vendor. Chips are hardware. VA Linux, whose intentions for the OSS world are unknown, could make a gesture of faith by setting up a factory to build these chips, and distribute them to members of the community. In my mind, this would help for me to take them as a serious partner of us Linux users.

    And it makes economic sense, too. Why buy your chips if you can make them? They could even be modified to be "LinuxChips", designed especially for running Linux and it's applications (KDE, etc). Since everyone would have them, it would make sense to want to buy more, and VA would be the sole supplier. A good deal, no?

    Lastly, since VA, being in the hardware field, doesn't do software development, we haven't had any free software from them, which is the typical currency here in the OSS community, which perhaps they did not know when they became a hardware vendor. To make up for this and appease Linux users, my idea as outlined above should be put into place. They are also strongly encouraged to support the community more, even if by a few simple web sites, or they may start to see serious drops in demand for their machines.

  • Re:Site slightly vapourous (Score:1)

    by Anonymous Coward
    The chips is real and has been tested. Perhaps you are thinking of the second core they are working on.

  • You could always (as I believe has been suggested here) place it on a FPGA. The price of FPGA development systems are quite cheap (we are talking low thousands I believe) - try VCC [vcc.com] for example.

    Using FPGAs to prototype and play with hardware is becoming more and more commonplace. I believe (though I must admit I've not seen hard evidance) that the sparc core will place and route on a Xilinx Virtex device.

    Remember though that compiling hardware is a lot more computationally expensive than software - need oodles of memory and CPU cycles :-)

    -- Michael

  • What's going on with Freedows these days, anyway?

    From the last status update (in September):

    Freedows is still in its early specification stages

    Sigh. After 4 years, they still haven't got any code out of the door. As ever, the project is all about beaurocracy and assigning resposibility to various projects, and and nothing to do with getting the coding done. I've been following the project since the early days, when it was still called Freedows '98 (remember that?), and it hasn't gone anywhere since then.

  • I'd like to refute the argument that there is such a large difference between hardware and software... As the article mentioned (if people read it), this is all implemented in an FPGA-- easily under $100 for a part. I use Altera FPGAs and for less than $70 I built a board using an SMT-to-through-hole adapter board with the FPGA and a setup for downloading designs. The software is free off of Altera's website. Much like traditional programs, I compile under Linux (using wine), and then try the program via a download to the FPGA. It's no different than having to buy a computer to write software; there is always a cost of entry, but once it's paid it is free from there on out. For other people to use or develop your "program", they need to make a similar outlay.
  • This is a great time. The same thing that happened to linux could happen here. If we have a free core that gets enough attention, then people will start poking at it. Pretty soon you start getting more and more people working on it. Eventually we will have as much time and talent as any microprocessor design company. I would hazard to say that there is more time and talent going into the linux operating system than any other OS. I say, build it! Make embedded processors! Build boards! Build a computer out of it and I will buy one! Binder
  • Looks like lots of people have questions for RMS, at least I know for
    myself I have plenty of things I'd like to ask him :) Why won't
    CmdrTaco run an interview with him? It'll be a blast...

    - Adi Stav


  • There is another group doing almost the same thing, they are the FCPU (Free CPU) people.

    Thanks to them, we no longer need to pay hefty amounts of money to get our hands on hardware cores anymore.

  • Not the manufacturing. Anybody with a billion dollars and 3rd world country w/o environmental laws can bake chips. It's about an open design and open documentation and no royalties or licencing for the specs.
  • ...ahhh, but the trick is working it backwards for the source material...

    I didn't say i could do it...
  • PS: I used to curse a lot on this site, but my firm now has a firewall on all email messages. Any of you young geniuses know how to get round it?

    see if it detetects curses with embedded tags:

    FU<I></I>CK = FUCK
  • You should go check and see how much Intel owns in VA Linux.

    You'd have better results shopping your design to a korean manufacturing firm. I bet that they'd jump all over that puppy.

    The "Big Boys" that own their own fab are too competitive and greedy to invest any help in this. You'd have to go out and sell your soul to lucifer.

    Pan
  • redhog writes:

    Strangely, though, [he] said that open-hardware was not that important, while the copying cost for hardware is quite high.

    Perhaps RMS misunderstood you? I certainly believe that RMS likes published specifications and documents about hardware. How else are we supposed to fiddle with it when we're writing operating systems and device drivers?

  • You should have checked the mailing list archive, the list is quite active and some work is done !


    The inherant problem of hardware is the delay between modification, edit--> Compile --> test, is much more important than in software.
    Today, with FPGA, we are able to do the cycle in less time but the testing of a hardware function is very time consuming


    Simulation of the first second of a CPU running a software can last several DAYS (On high end computer). The project is not vapourous it's delayed...

  • You have got to be on crack. Think about this reasonably. The initial investment to have a fab that will create a chip worth selling (or even giving away), is enormous. Truth be told, if you're going to fabricate chips, you have to go big. Really BIG. Otherwise, you're not going to make any profits (or even break even). That's why a lot of hardware companys like SUN don't bother with their own fabs, they sign agreements with companys like TI to do it for them.
  • less than 150 US dollars. check out altera and xilinx (with the obvious urls). they both have cheap development boards which can be used for this purpose. altera is the cheaptest, and the software is free.
  • If that can do better than nVidia then I am for it. I don't think anyone would play q3a on slow poke open video card...
  • I asked someone this the other day and couldn't get an answer. Its obviously not the song, the Greeks used essentially the same order (hence the work alphabet), but I am curious how and why the order was established. Some Slashdotter must know

    Skippy
  • I saw in the above posting that someone is working on free video cards. That would be great because:

    1. We're all sick and tired of paying $200-$300 for a glorified bit pump.

    2. It has traditionally been hard to get video card specs for writing drivers. Having knowledge of how each gate is wired would solve that problem (and punish the bull-headed, short-sighted vendors!).

    3. This is an area undergoing rapid development.
    Open/free development works well in this environment.

    4. Could borrow ideas (or at least the instruction set and software driver architecture) from Mesa and related free software projects. The potential synergies are tremendous!
  • Well, first thought is "what the fuck is the coffee machine doing on the UPS"

    Then I realized . . . what is the point of having the computers on, and no coffee!

    But . . . shouldn't the coffee machine have its *own* UPS ?
  • ... But you don't breadboard a CPU my friend. Have you ever heard of FPGAs? Go read about it. You can get one for as cheap as $200. I know someone that got one, designed his own 32bit RISC CPU with it, and built a motherboard on it, ported netbsd on it, and this was like 2 years ago. His cpu was clocking around 40mhz. Remember 2 years ago. Go read about FPGA.
  • When Andrew Cady said that this is just the design for a chip he is right. Right now you need an FPGA or a FAB to use the design. This is like a CD 15 years ago: couldn't make 'm yourself. OTOH right now you can, and I never imagined that. As I cannot imagine having my own FAB (well... it'll come). When you look at current technology, the next thing is Systems On Chip (SOC), and you need to get designs (copyrighted or not) to combine and make SOC's.
  • IMO, any product that relies highly upon knowledge is a definate candidate for "Open"-ness.

    The Open model fails as soon as money starts to interfere. (remember man-power is almost free in an Open world)
  • If you can get to a web site why don't you just use Hotmail etc for email?

  • My understanding and poor memory want to claim that hennesy and patterson's DLX architecture -- which (once again, from memory, so add salt) became concretised as the MIPS chip was the first RISC platform. Indeed, they could have patented (for once, something I conscider to be a patentable invention) the concept of constant complexity instructions, and prolly have made a bundle.

    But I guess they still made out ok.
  • Did anyone see the new job posted at Micro$oft? It was slashdot troll. Only duty is to blindly repost all of their rigged tests any time Linux comes up. Pays $25,000/yr.
  • All they are providing is a logical design.

    Having a logic netlist is only part of the battle when it comes to
    designing a microprocessor. It will be up to the user to do all of
    the physical design work. Since the wire delays between gates are
    becoming more significant than the gate delays, it can be argued that
    the physical design is the more difficult part of the problem. It is
    time consuming and expensive work, and needs to be repeated for each
    different fab.

    It's cool that they were able to load it into a xilinx part and get
    it working at 100 MHz. But a microprocessor inside a PLD will never
    be competitive in terms of cost or speed with a custom design.

    This may compete with the synthesizable core supplied by MIPS and
    others, but synthesizable cores are not the mainstay of their business.

    It is likely that MIPS and ARM will protect their intellectual
    property. MIPS probably hasn't acted yet because their lawyers
    haven't had time to digest this yet.
  • Actually I was talking about the Opencores processor.

  • If I read correctly the processor used a 106k gate fpga.

  • First off, the couch predates the patent system, so prior art kills that one form the door.

    The physical chip can be purchased. It is what is known as a field programmable gate array. Essentially a "field" of gates (transistors). With the development software you can organize (program) the gates to do specific things [newscientist.com].

    Here are other fpga processor articles:
    Article [slashdot.org] Article [slashdot.org]

  • I believe you're talking about a barrel shifter. Interestingly enough, you probably have greater need for one in your video accelerator than in your CPU proper. (Cracking RC5 on your video card, what a concept!)
    --
  • FPGA's are a possiblity for hobbyists.
  • "The approach is similar to the practice among amateur programmers of making software utilities and games available free over the Internet." I guess to be a pro programmer you have to sell advertising space on your laptop, do a commercial spot for Pepsi, and refuse to code for at least one whole season while renegotiating your contract.
  • The article _said_ it wasn't going to be released for a few days. I believe the designer is taking finals this week. Cut him a little slack. You can wait a few extra days.

    In hardware it's a little more important that shipping code be bug free.

  • The only money VA Linux has is their stock value, and I don't think IBM will accept stock for a fab run.

    And... /. did not acquire VA Linux. VA Linux acquired /. There is a BIG difference.

  • The OpenRISC 1000 core preliminary ISA spec lists a variable bit shift operation, in both left & right, signed & unsigned.

    Of course, simulated logic isn't your best place to do RC5 cracking, even if it is hardware simulated (programable logic).

  • Please can I request that the lovely instruction in a Pentium that does the n-bit shift in one clock cycle be in the core. Just so as I can crack RC5 quicker ...
  • Once a chip is made, its made. And AFAIK, most chips don't really have any problems, or at least problems which cannot be resolved.

    Depends. A lot of your industrial controls and such use FPGA's now, meaning bug fixes, firmware upgrades and such can be be re-applied to the existing hardware, usually by just hooking it up to a laptop for a few minutes. It's more limited than software bug fixing, but not as much as you might think.

    Of course, if you're talking custom IC's, then "bug fix" generally involves an incinerator for the old hardware and a purchase order for the new hardware.

  • I attribute it to software engineering and software engineering being two different art forms entirely. The former being a more 'hypothetical' art form, becouse it's simply the directing of flows of information, and is much more of an 'art form'.

    Hardware Engineering, on the other hand, is much more in the physical. It is made into something you can feel and see.

    Cars shouldn't be free. But the idea of an internal combustion engine, connected to a device with 4 wheels, should be free.
  • Even though I believe that the firm itself got its advice from some punky West-Coast firm, it is firmly in the grand legal tradition of IP law invented over the last two hundred years, ever since the writing of the US Constitution itself, the first document to introduce the concept of a civil patent.

    Crap. Patents are not an American invention - check out the English Monopolies Act (Statute of Monopolies) 1624. Other things otherwise smart people regularly claim are American inventions:

    • Trial by jury
    • Representative government
    • Common law
    • Automobiles
    • Railroads
    • Radio
    • The English language (I kid you not - this got me into a foul argument with a bank teller who checked my passport when cashing a cheque - "You speak good English", "Well, yeah, I'm from England", "Oh, do they speak English there?")
    • Computers

    Nick

  • 1. We're all sick and tired of paying $200-$300 for a glorified bit pump.

    The last video card I bought was a S3-based one from a Chinese brand of some sort. I picked it up at a computer swap meet a few years ago for under A$100 (that's about US$62). Surely such lowest-common-denominator (though still accelerated to some extent, I believe) cards still exist, don't they?
  • Let me know when I can download nanoassembler source files for this kind of stuff.
  • This story made me think of the good ol' days on /. when they would have stories about over-hyped pipe dreams like the Freedom CPU Project (http://f-cpu.tux.org/) and Freedows (bad name, neat idea).

    Can we have more stuff like that? That was fun. What's going on with Freedows these days, anyway?

    --

  • though you can use them to simulate code.

    Students write complete Verilog descriptions of the processor, which can by synthesized on an FPGA or sent off to a fab to make "real" chips.

    ARM distributed simulators so people can verify that their code runs on the chip without actually having one.
  • If I read correctly the processor used a 106k gate fpga.

    Not a chance. Those didn't exist in the late 80s/very early 90s. It is a custom gate array (maybe not full custom). I have one in my basement. It might not have 20k gates, it might have 40k. I don't know how to tell from the packaging.

    I could beleve that the SPARC in the SPARCStation1 (~15MIPS? ~20? I think 20, I'm almost positave it was a 20Mhz CPU) was 106k gates. Unless it had an on-chip cache, those eat gates like no tommrow (4T or 6T per bit of SRAM is state of the art!).

    Remember I'm talking the very first SPARC CPU sun sold outside the labs. Back when $10,000 got you a ~4MIPS VAX. Back before there was a "SPECmark". Back when RISC was "no multiply instruction, branch delay slot, one instruction per cycle, god that's fast!". This was the very second (I think the IBM RT/ROMP made it to market first) RISC CPU in the (Unix Workstation) market.

    Still 104k gates can be done on a ~$10 gate array, subject to routing constraints and grouping constraints.

    Another CPU I'm pretty sure of the gate count for is the original 68000, roumor has it it used almost eaxactly 68000 gates. So that is also in shooting range for a dirt cheap FPGA. Of corse the 68000 is dirt cheep these days too, the 68000-decendent in the PalmPilot has lots of integrated perphrials (LCD controler, RAM controler, and so on) and it costs under $20 (for the DragonBallEZ; the DragonBallVX in the Palm IIIc is more like just about $30).

  • Those who are reading are characteristically always forgetting to guard their own mother to grab it themselves. Ruler of creeper, mortal, and scallop, this is a machine is a kludge, after all, as "wanting." it is not generally understood by less advanced life forms, and they'll show a commercial for a while.

    Cool! Mark V. Shaney reads Slashdot!

  • It was no more radical than Acorn's guys coming up with the ARM and RISC architecture, which completely blew the lid off the CISC market.

    Did ARM antedate, say, MIPS and SPARC (which I think may have been the first commercial "merchant semiconductor" RISC microprocessors)? (I suspect it didn't antedate, say, the IBM 801 research machine, or the Stanford MIPS or Berkeley RISC projects.)

    Acorn may have had one of the first, and perhaps the first, personal computer with a RISC processor, however.

  • What does anybody know where ELF format stands at in the GNU GPL scheme of things.

    "Stands" in what sense? ELF was invented by AT&T, as the object file/executable file/shared library/core dump format for System V Release 4, and was published in various SVR4 manuals; I don't think they made any patent or other claims limiting implementations of it (or, if they did, they apparently didn't get too far, as the GNU toolchain supports it and Linux and the BSDs use it).

    None of that would, as far as I know, preclude Ricoh from patenting an extension to it.

  • Please can I request that the lovely instruction in a Pentium that does the n-bit shift in one clock cycle be in the core.

    The instruction in question does so because, as another poster noted, the Pentium in question has a barrel shifter (or another circuit capable of shifting n bits in one clock cycle), not because the instruction is defined in the instruction set architecture to be a one-clock-cycle shift.

    I.e., what you meant to say was "Please can I request that the core have a barrel shifter so that shift instructions can shift n bits in one clock cycle?"

  • CPU designs have been efectively free for a long time now, and that hasn't changed much. Ross was the last company to seriously try to capitalize on this idea

    "Capitalize on this idea" in what sense?

  • Well - that IS a problem, but there are solutions available. First, there are open source software solutions that are being developed to do gate level synthesis. The Alliance VHDL environment can already do Xilinx synth as an example. The place and route still requires commercial solutions, but there are folks that have access to those tools around ;-)

    Steve
  • Yes, or you can use the software ARMulator written by ARM Ltd, a version of which is GPLed. It might be a bit slower though :-)
  • "and if you people haven't heard, that's the way the wind's blowing."

    remember "pen computing"?

    Sometimes the wind blows. Sometimes it just breaks.
  • Who said free software, free hardware, or free anything, was about money alone?

    There are more potential benefits, than money. There is the hope, and goal, of a better design, through better understanding, and more eyeballs finding problems.

    Also, a more open hardware design, can lead to better software written for it, especially compilers.
  • Recently, I was listening to RMS, and after the show I was able to ask him some questions. One question I asked was his opinion about where the border between hardware and software is (firmware comes to mind), and if he would like to see an open-hardware movement. Strangely, though, said that open-hardware was not that important, while the copying cost for hardware is quite high. Still, I beleave such a movement is important, and partly, the PC is such a standard. Perheaps time to start another open-movement, and to convince RMS that he is wrong :)
    --The knowledge that you are an idiot, is what distinguishes you from one.
  • It is good that this type of thing is available, for the same reason OSS is good.. and free information is good.
    It will not, however, 'undermine' commercial vendors. At least, not those who have products with any value.

    In my view, OSS sets a baseline for what is acceptable. Look at sendmail.com. They build management systems around sendmail, and they make money. They add value to sendmail. They do not make their money because they control the ability to send mail.

    A standard software practice by MS is the feared embrace, extend, extinguish. Here's a (barely) hypothetical situation. (I'm making this up, but it's probably somewhat true).

    MS releases a DNS server, and a nicely integrated management system for administering it. Now, provided it's stable, and follows standards, if the front-end is good enough, it may be beneficial to me to use it. My reason for using it is not because I need DNS service, but because I feel it is easier to manage than my current DNS server. Then, after configuring my whole zillion dollar enterprise to use it, I find out that it uses some kind of proprietary record types.. and there is no way for me to convert back to standard BIND config files. This happens to others, and eventually, if you don't support their version of things, you are out of the market.
    So, now people use it because they simply need DNS service, and they have nowhere else practical to turn.
    MS will insist people use the product because of it's superior interface, when really poeple use the product because they no longer have a rational choice.

    THIS is the practice that OSS puts a halt to. You make something GOOD? Great.. people will buy it. You try to use the fact that your product is popular to subvert the baseline to control the market? OSS will stop you.

    If a bunch of people can get together and make something that really is a contender up against ARM and other chips.... and it costs an arm and a leg to license ARM.. well... perhaps ARM really isn't worth it. On the other hand, if the ARM core is technically and functionally superior.. they have nothing to worry about..
  • These things should be included with the story...

    http://opencores.itvc.com/ [itvc.com]
  • They say on their webpage that they need verilog coders as well.
  • perhaps not completely ideal, but you can synthesize one of these cores and drop it into an FGPA development board. This doesn't cost much and is a standard project in Computer Engineering type programs.
  • This is non-news if I ever heard it. CPU cores and even complete CPU designs have been free or near-free for years.

    Sun started giving away SPARC uP designs last year - before that they were charging a stunning $99 for them.

    Oh, there is the catch that in order to actually build the chips, you'll need a modern stae-of-the-art semiconductor fab facility, which should run you about US$3e9. Or you could spring for a few tens or hundreds of millions and buy a slice of some else's fab, if they have excess capacity. (I don't expect there's much of that around right now...)

    CPU designs have been efectively free for a long time now, and that hasn't changed much. Ross was the last company to seriously try to capitalize on this idea and look where it got them!
  • Actually, radio *was* invented here. This was settled legally in a sensational lawsuit (which was appealed IIRC to or near the Supreme Court) between Nicola Tesla (who really invented it first), and Gulielmo Marconi (who independently invented it a short time later and proceeded to devlop it much faster than Tesla.)

    Although Tesla was not an American by birth, he got here as fast as he could!
  • What is the workers don't make products? What then? You're forgetting the teachers, lawyers, delivery men, accountants, and everyone else in the service sector (which is more than half of the US economy, btw).

    Ryan
  • A friend of mine runs http://www.free-ip.com [free-ip.com]. Free-ip is a lot less hype, and a lot more content. Several cores are online now, and there is some exciting new stuff ready but not yet released. Obligatory grumble, I've been trying to get Slashdot to run a link about it for months, but it never gets approved. ;)
  • John S Montoya.
    You killed my cash cow.
    Prepare to try.
    --
  • Won't be long before we figure out how to do to your field what we've done to software and are doing to hardware now. I really can't think of a more deserving group.
  • So we're supposed to believe that since companies don't have to pay for the core design, a lot of small companies will be building fabrication plants and making their own chips?

    If a company can afford to MAKE chips, I imagine the license is the cheapest thing they ever buy.

    Esperandi
    This is like saying we could get more poor people on the net if we bought them a $10 keyboard. Surely they can do the rest?
  • circu.its.tudelft.nl [tudelft.nl] is a site about Open Hardware, too, but this is more of a research point of view.

    It's pretty old, but contains interesting thoughts about Open Design Circuits as they call it, and has a number of good links [tudelft.nl] to other sites about Open Design/Open Hardware

  • What the hell are you talking about? THIS IS JUST SOFTWARE. This is the chip DESIGN, not the ACTUAL PHYSICAL CHIP. This is just software like any other, except instead of you creating the hardware on your home computer by copying it into RAM, someone will have to physically manufacture it. And guess what, the open hardware works in many (most) industries. Look at the common couch. Do you think that there is a patent on it? Of course not. Yet plenty of businesses make money on furniture. You are totally misinformed. Try reading the article.
  • Unfortunately, they're not offering a 32-bit core, they're starting work on one. Completion may be some time off.

    Bear in mind that low-end Intel architecture CPU parts start below $50. There's not much maneuvering room here.

  • Check out http://geda.seul.org for Gnu EDA. There is a VHDL simulator, Savant, which I've worked on. More info at http://www.ececs.uc.edu/~paw/savant. There's a Verilog simulator at http://www.icarus.com, too. All of these projects could use help from interested parties!
  • You might have noticed, the post was moderated as "funny"? IT'S A JOKE. Maybe a funny one, maybe not, but the author is not misinformed.

    Apparently you're not the only one who didn't get it, though; some other insightless person actually moderated you as insightful.

  • This is great, all I need now is a free mp3 decoder core and a free 802.11 core, and I would have a hell of a mp3 player.....

    Or a beowolf cluster based on these things.....

    Grtz, Jeroen

  • OK, so I know that this doesn't mean we can all start stamping out OpenRISC processors in our basements, but the idea of "Open-source Hardware" is still very powerful. It creates an entirely new species of hacker: the hobbyist chip engineer. Of course, you won't always be able to test your neat hardware designs (or could you? very low level emulation?), but for those who have access to fab plants (I know you're out there), this could be a very fun project.

    Also, this further commoditizes the chip market. I think we've all benefited from the feirce competition between Intel and AMD, considering that things like sub-$1000 computers, SIMD instructions in end user chips, and of course dramtatically lower prices on x86 procs wouldn't be here yet if not for this rivalry. Still, to compete with Intel, AMD had to do huge ammounts of internal engineering to duplicate the functionality of Intel's chips. Not so with OpenRISC. Instead of wasting time cloning other developers' features, engineers will simply add their own, speeding up the advance of the design much more. Prices will also be considerably lower due to (duh) lower R&D costs. It like the difference between NT and Solaris competing against each other, versus the Linux distros cooperating for mutual benefit.

    One question, though, is the OpenRISC specification going to be released GNU-style (no proprietary extensions allowed, ever) or BSD-style (just give us credit and you can fragment it like a FAT partition with your corporate crap)?

    P.S. Don't get me wrong, I love the BSD's and use them daily, but I think the GPL is a much better license.
  • I posted a link to Open Design Circuits [tudelft.nl] a while back. Maybe these guys should get together?
  • > the company takes "the protection of our
    > intellectual property very seriously but can't
    > comment on specific cases. We prefer an open
    > partnership model that broadens access to ARM
    > architectures."

    Translation: To crush our enemies, to see them driven before us, and to hear the lamentations of their women.

    The second sentence has no semantic content, therefore no translation can be offered.

  • bmfosh (Score:2)

    by gmp ( 155289 )
    I wrote a paper last week on this topic. For those two think this idea can't work, think again. The ecconomic principles involved are similar to those involved in open source software, and once a design is released, it is out there forever, free as the air to common use. Does the computer industry need OSH the same way it needs OSS? I think it does.

    "Business Models for Open Source Hardware [nyu.edu]"

    ---

    gmp

  • IMHO, if this project gets truly off the ground, and stays aloft, we could see some radical changes in the way processors are designed.

    A sweeping statement? Pah! It was no more radical than Acorn's guys coming up with the ARM and RISC architecture, which completely blew the lid off the CISC market.

    If these OpenCore folks, or any contributer thereof, can come up with a new way of doing things that out-performs the old (very likely, IMHO), you'll see it in "mainstream" chips within 4-5 years, even if that means some companies scrapping designs of their own.

    Chip manufacturers may be very protective of their so-called Intellectual Property, but give them ten seconds and a decent spec, they've absolutely no hesitation on copying the better ideas off other companies.

    In a way, that's good, as it means the OpenCore group don't have quite so much of the "Not Invented Here" syndrome to fight. On the other hand, it does mean they can expect a lot more licence violations and other interesting distractions.

  • Damm, it's happening already (Score:3)

    by SurfsUp ( 11523 ) on Tuesday February 29, 2000 @05:24AM (#1238155)
    The logical next step after open software is open hardware. The only reason for a lag is: hardware simply has higher capital cost than software. Until now, that is - when open tools bring about the demise of proprietary hardware design tools, the cost of designing hardware will drop dramatically, and we haven't seen nothin' yet. The cost of actually creating silicon is going to drop like a rock too.

    The only real suprise is to see it happening already, before the open source hardware development tools have gotten to anything like maturity.

    Open dvd drive, anybody?

  • Not gonna work (Score:3)

    by funkman ( 13736 ) on Tuesday February 29, 2000 @06:28AM (#1238156)
    Hate to be a pessimist but there are some reasons this will fail.

    First, people working on this in their free time time will not match the quality or technology of a firm which does do chip design. Remember that fixing and rereleasing buggy code is much easy than fixing a chip. Once a chip is made, its made. And AFAIK, most chips don't really have any problems, or at least problems which cannot be resolved.

    There are tons of chip patents out there waiting to get stepped on. To get around these hurdles will be a nightmare because there is no way to disprove you are violating a patent. (unless you can prove that patent is invalid)

    The technology curve will be lagging compared to chip firms. This point could be voided if a grad school(s) took this project under its wing. I believe Carnegie Mellon takes care of MACH and does tons of research in trying to make it better. But I think a lot of that research is sponsored by companies who may get some rights to the technology discovered. (any know those issues?) This could be a good starting point for grad school students to try new designs on a chip to see what happens.

  • RMS is a flake; free interfaces, not source! (Score:3)

    by TheDullBlade ( 28998 ) on Tuesday February 29, 2000 @03:32PM (#1238157)
    As a philosopher he's a joke. His views are not self-consistent. The whole free software thing basically comes from a pleasant time in his life when all the software he saw was software he could hack. His whole justification is basically "that was nice, let's do that." He has other rationalizations, but they are recognizable as such by the way they break down on logical examination.

    If he hadn't come along, Linux would have been another free BSD variant. The people who put their work into GNU projects would have put their work into other free projects and we'd have more or less the same stuff.

    All in all, I think we would have been better off without him. He's an annoying nut who makes the rest of us look crazy and is always trying to steal the spotlight and tell us how most programmers should be broke like most musicians (go read his stuff, if you think I'm making that up). We could live without the GPL, and we could certainly live without Eight Megs And Constantly Swapping!

    That source code be free is not so important. It's nice, but not crucial. What is important is that standards be open. It is essential that anyone willing to invest the effort (and smart enough to manage the task) be able to build their own replacement for any one part. Secret instruction sets are wrong, patented reading mechanisms are wrong, proprietary interfaces of all types are wrong.

    The problem which destroys freedom is not that we can't see the internals of others' mechanisms, but that we aren't even able to produce devices which connect to the interface visible on the surface, or to produce fully compatible replacements for these proprietary devices.

    For me, the obvious example of this is the gaming console. It is absolutely unacceptable that the producers of the console have control over who produces games and peripherals for the console. Sure, you'd have to pay more for your new Nintendo toy, but you'd save the difference in the cost of games, plus you would have the freedom to be your own censor, rather than allowing Nintendo to decide what games are good for their company image, and therefore acceptable for you to play.

    The importance of open interfaces should be written into all IP laws. If all interfaces are open, then proprietary and open devices can coexist and free market dynamics assert themselves over the strange positive-feedback dynamics of gatekeeper marketing strategies reinforced by outdated IP laws.

  • Open Hardware/Software Branding (Score:3)

    by G27 Radio ( 78394 ) on Tuesday February 29, 2000 @08:28AM (#1238158)
    I'd really like to see an organization that performs some kind of verification of the "openess" of a product whether it is hardware or software. They could allow products that are open to bear their logo.

    Initally it would be nice because it would make it easier to identify products that conform to an "open" standard. Later on, as open products become more common and prove themselves to the public, it may encourage hardware and software producers to keep their products open.

    This assumes that it will eventually become a more widely known fact that "open" products are by definition better. Of course, the logo may raise awareness in the meantime.

    Really this is just an idea that's been bouncing around my head. I'm not convinced that it would be worthwhile at this point, but it seems to have possibilities. This article sort of reminded me to mention the idea.

    numb

  • Re:free as in hardware? not free beer (Score:3)

    by Esperandi ( 87863 ) on Tuesday February 29, 2000 @05:12AM (#1238159)
    Actually, a company is starting a sort of free-hardware-for-free-code exchange program. If you've got extra hardware and wanna "give back", send it to the company. Then they wait until an open source project comes around that can use it and they give it to the programmers.

    It's really a very interesting concept, there was an article in Wired News about it a month or so ago (I submitted, Slashdot rejected)

    Esperandi

  • Just a reminder (Score:3)

    by mr ( 88570 ) on Tuesday February 29, 2000 @05:27AM (#1238160)
    That, when Mr. Metcalfe suggested the Transmeta processor should be OpenSourced, a group of people said that hardware can't be OpenSourced because hardware can't be OpenSourced.

    As others pointed out then, yes it can.

  • Open letter to Microsoft (Score:3)

    by MattMann ( 102516 ) on Tuesday February 29, 2000 @08:12AM (#1238161)
    People like you are going to keep the open source movement from advancing.

    Dear Microsoft,

    I wield a lot of influence. Apparently, I even have the power to keep the open source movement from advancing. Please cut me a big check and I will this little problem of yours go away.

    Sincerely Yours,
    Matthew Mann

  • RMS confuses me sometimes... (Score:3)

    by Skald ( 140034 ) on Tuesday February 29, 2000 @09:02AM (#1238162)
    One question I asked was his opinion about where the border between hardware and software is (firmware comes to mind), and if he would like to see an open-hardware movement. Strangely, though, said that open-hardware was not that important, while the copying cost for hardware is quite high.

    I have read Mr. Stallman's opinions for some time now, with an eye towards piecing them together and finding the ideas which underpin them. It ain't easy. For one instance of many: he often talks as if a particular principle were important for its own sake; then later you'll see him discount the importance of something similar on purely practical grounds. The difficulty of copying hardware is a good example.

    I don't mean to imply that his views aren't internally consistent; he's not a dumb guy, after all. But they can seem paradoxical at times.

    Part of the problem, I think, is that he's not very forthcoming with the personal views which inform his public stances. This is entirely understandable; he surely doesn't want his personal views to be a liability for positions he wants to promote.

    Nevertheless, he has set himself up as a public philosopher, and this causes confusion. He mentioned somewhere-or-other, for instance, that he's an atheist... yet he frequently speaks of rights and ethical issues as if they were objective, rather than subjective. Not a self-contradictory stance, but an unusual one which begs questions.

    I was puzzled to find him lukewarm of free hardware designs too. It'd be nice to get a chance to interrogate him on his deeper worldview sometime. One might even be able to make a case that he owes us an explanation. :-)

  • SPARC already has done it (Score:4)

    by bjb ( 3050 ) on Tuesday February 29, 2000 @05:02AM (#1238163) Homepage Journal
    A few months ago, we already read about a similar situation with SPARC processors [slashdot.org]. Granted Sun released an older version of the SPARC processor and they also released it under their SCSL license, but nonetheless..

    I think it's cool that they release this kind of information. Unfortunately, who the heck has the resources to punch out one of these things besides the major chip manufacturers anyway? Maybe if someone was really bored and had a very large bread board (and a lot of spare time and cash on their hands) they could brew one up out of ICs and other solid state components. Probably run hotter than a docker's arm pit and use more energy than you can shake a stick at. Interesting, regardless.

    --

  • Re:SPARC already has done it (Score:4)

    by stripes ( 3681 ) on Tuesday February 29, 2000 @06:03AM (#1238164) Homepage Journal
    Unfortunately, who the heck has the resources to punch out one of these things besides the major chip manufacturers anyway?

    For a MicroSPARC? I would guess that you need full custom logic, which isn't that expensave if you are buying tens of thousands of the device. It's suck if you want, say, three.

    For the OpenCore CPU? That one is being targeted at FPGAs, which cost a small number of thousands (like two-ish) allready soddered on a PCI bord and ready for love (this would be for the million-gate-ish FPGAs). The "smaller" FPGAs, like 10k to 200k gates are available for way cheep, like under $10 for some of them (with 100k gates even!), very few over $100.

    Remember that the first SPARC implmentation was a 20K gate array, it ran at about 10MIPS if the wind was going it's way (Sun 4/110). So you can probbably design a low speed CPU on a $5 FPGA and have room for a USB interface. Spend $40 and you have room for some periphrials too.

    Unfortunitly the hard part is the dev tools. Many cost over $100k. I don't know of any Open Source FPGA design tools. Wish I did. I would love to design a non VHDL language!

    For FPGA examples and prices look at xilinx [xilinx.com]. Not sure where to go to look for software pricing.

  • This is nice, but... (Score:4)

    by GregWebb ( 26123 ) on Tuesday February 29, 2000 @06:40AM (#1238165)
    EETimes has a story about how OpenCores is offering a free 32-bit processor IP core in a move that could undermine such commercial IP licensors as ARM and MIPS.

    I'm delighted to hear that someone's actually making headway on this sort of thing, but...

    That headline sounds dangerously like undermining ARM and MIPS is regarded as a positive acheivement because what they're doing is fundamentally evil. It isn't, and posting that sort of thing on the front page of slashdot only helps alienate business from the open development community. If we're going to succeed, we need their help sometimes.

    Greg

  • Software for designing opencores? (Score:4)

    by gargle ( 97883 ) on Tuesday February 29, 2000 @11:40AM (#1238166) Homepage
    Is there good free software to go along with the "opencore"?. i.e. Verilog simulators, Xilinx/Altera style digital design software,etc.? These software packages are quite expensive, and without good free software equivalents to go along with the free cpu designs, the whole thing is pretty much pointless for the hobbyist.

  • free as in hardware? (Score:4)

    by MattMann ( 102516 ) on Tuesday February 29, 2000 @04:59AM (#1238167)
    OK, we won that whole open software thing. Hardware? Time to start stressing the importance of "free, as in beer" :) heh heh

  • A lawyer comments (Score:5)

    by Anonymous Coward on Tuesday February 29, 2000 @05:09AM (#1238168)
    As a practicing attorney at a leading IP firm, may I be the first to congratulate these kids on what they have achieved? To liberate intellectual property from other people, to destroy the value of their IP and to do it all wholly legally is truly an inspiring achievement.

    Could I perhaps, however, blow the trumpet for my own profession and point out that the entire IP edifice which made this possible was the creation of lawyers? It was lawyers who created the statute and precedent which makes a viable IP solution to the issues raised by the GPL and Opencores possible. Without dedicated IP lawyers, Opencores would be mired in a sea of uncertain intellectual property rights. Even though I believe that the firm itself got its advice from some punky West-Coast firm, it is firmly in the grand legal tradition of IP law invented over the last two hundred years, ever since the writing of the US Constitution itself, the first document to introduce the concept of a civil patent.

    Lawyers. You don't have to love us, but we gave you IP in all its glory.

    John S Montoya, formerly known as streetlawyer.

    PS: I used to curse a lot on this site, but my firm now has a firewall on all email messages. Any of you young geniuses know how to get round it?

  • Re:free as in hardware? not free beer (Score:5)

    by dattaway ( 3088 ) on Tuesday February 29, 2000 @05:43AM (#1238169) Homepage Journal
    People like you are going to keep the open source movement from advancing.

    Don't be so paranoid. There's a reason why open source became adopted by many people.

    Take a trip back to 1997. Windows was so buggy and it wasn't just me. Everyone else was losing data and lost time. Remember when NT was released? NT 3.5? If that wasn't enough, remember NT 4.0? At work too. Viruses were rampant. I was afraid to look at someone else's computer for fear it would break out and catch something. My computer was in isolation and its inbred closed operating system still crashed. It was the shits.

    Only because many of the show stopping bugs have been found in Windows that many people are able to find it completely tolerable. I won't return to Windows, because now I see its a closed source trap.

    Open source has gotten a lot of attention lately not because it is economically free, but because companies/people are profitting off of it.

    People are making a profit off it? Great; however, I make my money doing something else, but when computers crash, it makes my job harder. That's why I enjoy the stability and open nature of free software.

    Open hardware design is great for people like me who often find a need to work with existing machines and adapt them to the changing needs we have. Open things open up opportunities. If they close them, find another line of work.

  • hehe, I just thought of something (Score:5)

    by aheitner ( 3273 ) on Tuesday February 29, 2000 @07:05AM (#1238170)
    If you want a Free implementation of an ARM-compatible CPU, just turn to anyone in the CMU ECE dept. who has taken 18-347 (the 2nd level processor-design course). The semester-long project in that course is to implement an ARM-compatible CPU in Verilog. Structural in the fall, procedural in the spring :)

    They don't synthesize them (no real point), but they run them in simulation, and you can write assembly or use the GNU ARM compiler to make code for it. I believe (haven't taken the course yet) they give you a few of the more complex parts to save time (I think they give you the FPU). You write the rest in your project groups.

    Just have to talk someone in the class into releasing their code under a Free license. Have to check about any restrictions on any code they were given to use, tho

  • Believe it or not... (Score:5)

    by aheitner ( 3273 ) on Tuesday February 29, 2000 @05:22AM (#1238171)
    It's very easy to fab chips these days. Lots of companies are out there doing custom chips for various applications (ASICs, application-specific integrated circuits). All of those companies have a premium on fast design, and getting the hardware right on the first fab run (since that's a very expensive and time consuming step). And all of these companies use outside fabs to actually make the chips.

    So you can take these OpenRISC designs over to your favorite fab company, and just have a bunch made. Motorola does this (for pretty big orders). There are quite a few others. It's not even that expensive; professors here at CMU get their experimental designs fabbed (in small quantities) when they're ready.

    But what's even cooler is to get a nice Xilinx or Altera FPGA, and just download the chip on there. That's how the designers work. It's not the most price effective (a good enough FPGA board will cost you a couple k, and you need the synthesis software, and a good copy of Verilog or VHDL) but it's way more fun, since you can play with the chip yourself and change it around. We did this (with a pretty simple chip design) in my 18-240 class.

    What's more important is that there's plenty of competition for fabbing chips, so the market is good for consumers (rather unlike the traditional chip market). So you're likely to get a considerably better deal. You can't get the absolute latest .18 micron copper process (unless you wann pay IBM to fab for you, and I bet they're steep). But lots of guys have pretty good .25u stuff. Band together with some friends and get an order going!

    [sigh] too bad the OpenRISC is all VHDL and I only know Verilog, or I'd take it over to the lab here and see if I could get it on the university's FPGAs :)

    But this could end up really changing the way hardware is made and sold. Remember, there are a lot of university ECE departments cranking out good work, and looking for research projects to work on (check out the OpenRISC 2000 on the OpenCores website). The origin of all of UNIX, and more recently of Linux, was people at university CS depts. looking for fun stuff to hack on... all it takes is projects in CVS for the collaboration to begin...

  • Site slightly vapourous (Score:5)

    by warmcat ( 3545 ) on Tuesday February 29, 2000 @05:29AM (#1238172)
    I read this article yesterday and was pleased to see a downloadable 6502 in VHDL and DES encoder on another linked site from the article (http://www.free-ip.com/).

    The site that got the most discussion in the article about the RISC core seemed to me to be mostly vapour. Click those links on the left of the site (eagerly reproduced in an earlier post) and you'll see most of them are in the pre-discussion phase, let alone coding.

    Bruce Perens or somesuch said it best a few months ago, put out your code first before the vapour.

    -Andy

  • Ever since I started using free software (5 years ago?), I've seen people on the net starting apparently impossible projects. You say to yourself, naah, they'll never make it. FYI: I thought that when I heard about The Gimp for the first time. Imagine my surprise when I installed the RPM 2 years later. I thought that when I heard about WorldForge last year on slashdot. I thought that about opencores -- they have a working chip now. I thought that about Freedows -- hey, they failed.

    Overall, you've gotta respect people with grandiose ideas. They are the ones who achieve the greatest things. Or may not. One thing is sure, if you're not slightly megalomaniac, you can't do it, or even understand it.

  • This is soooo cool (Score:5)

    by SgtPepper ( 5548 ) on Tuesday February 29, 2000 @05:07AM (#1238174)
    First Phase; to design and to manufacture a prototype of a computer system in a form of a PCI card that will run GNU/Linux:

    we define a design and verification methodologies
    we design a library of small reusable components such as adders, multipliers etc
    we design a basic set of IP cores (microprocessor, PCI and SDRAM controller and communication controllers)
    we design and manufacture a prototype board with several hundred thousand gates FPGAs
    we support our microprocessor architecture with a port of GNU C Compiler, GNU Binutils, GNU Debugger and
    architectural/implementation simulator
    we support our microprocessor architecture with a port of Linux kernel


    This is a project that should have been started a looooooong time ago. Of course the monetary barrier of entry was quite high, but that's all changed....next we'll have LinBoard with the LinProcessor....ooooooh, and all open sourced...

    also i just wanted to point out that this is /much/ more then JUST Processor cores...check out this list of stuff:

    Comm Controllers
    ATM AALx
    Eth MAC 10Mbps
    FireWire
    IP Engine
    Serial UART
    USB Controller

    DSP Cores
    FIR Filter
    MAC unit

    Microprocessors
    OpenRISC 1000
    OpenRISC 2000

    System Controllers
    ATA-3 (EIDE)
    PCI 32 Bridge
    PCI 64 Bridge
    SDRAM Controller

    Video Controllers
    CRT (VGA)
    LCD

    Prototype Boards
    Micro FPGA board
    Super FPGA board

    now granted some of that is "wanted", but some of it is there....this IS cool tech :)

  • The argument against (Score:5)

    by Some Strange Guy ( 106728 ) on Tuesday February 29, 2000 @09:07AM (#1238175)
    I believe there are a number of reasons free hardware just won't work like free software. I could be (and actually, hope I am) wrong, but here goes:

    1. Chips don't have a near-zero replication cost

      • With software, 99+% of the cost lies in the development and documentation of a software package. Distribution is virutally free, and the economies of scale are pretty ideal. This is not the case in hardware, especially on high-end chips; Many of these chips cost $50-$150 each to fabricate, when die size, yield, package, etc. are factored in. So the development costs for leading edge technology is an extremely small chunk of the overall cost of the chip.

      • Compare this to software, where the bandwidth cost of transmitting a new linux kernel tarball or the presseing cost of a distro CD is pennies or less, and that's a very significant difference. The lack of a significant replication cost is a key reason free software works; there's no real monetary risk involved in distributing CD's.

    2. Chip design is not terribly modular

      • Yes, you can seperate the areas of a chip out into different boxes and define interfaces, etc., but even the best writting verilog is still fairly spaghetti-like. There are global signals that run everywhere, small changes can affect timing in many, many different areas, and it generally takes a pretty good knowledge of an entire chip to be able to make modifications without breaking things

      • Contrast this with (well-designed) software, in which there are generally interfaces and black boxes. In a free software project, I can usually go in and make edits relatively easily; the assumption that changing the internals of a black box without changing an interface won't break anything generally holds true. This is an important point that makes improvement by the masses feasible.

        In every verilog or vhdl-based project on which I've worked, a small group of coordinated, very closely communicating engineers was critical to ending up with a good design.

    3. Circuit design has a much steeper learning curve

      • To create a good VLSI project, you need to have quite a bit of training/experience under your belt. It's relatively easy to start programming and gradually learn what works and what doesn't. As a result, the pool of available programming talent dwarfs the pool of available circuit design talent.

    4. Revisions are expensive

      • Referring back to ESR [tuxedo.org]'s C&B, releasing often is critical to the success of a free software project. Hardware doesn't afford that luxury, especially chip design. You can see the results in simulation, and possibly in some sort of FPGA emulation rapidly, but the end results are on VERY long term cycles. Software allows you to incorporate changes and include them in the end release product in minutes. Chip revisions cost thousands and thousands of dollars for new mask sets, which make frequent revisions pretty much impossibly for even the big corporations, much less the little guy.

    5. Moore's law hurts

      • There are two major spectrums to address here.

      • For high end chips (Your Athlons, Coppermines, Alphas, etc.), the design deadline is absolutely critical. The luxury of "Being done when it's done" that is a central tenant of free software development doesn't work in chip design. There's a window of opportunity for a fabrication process that you have to hit, otherwise you'll either miss the process completely, or you'll be slower and more expensive than someone that did hit it. And what makes free software really cool is that it's often higher quality than the closed equivalents.

        The other major area of chip design is embedded processor design. Here the argument is very similar, but performance/cost is king. Competing here is even harder, because the demand curve for these kinds of devices is so nearly horizontal that any cost increase really hurts.

    6. Product lifetimes are generally short

      • This is related to the Moore's law point. Look at the linux kernel. It has been under active development for upwards of 10 years now, and it's just in the last year started to make a significant dent in the marketplace. With a hardware cycle, you typically have to scrap what you have at the moment every few years and start over, just to take advantage of the improving fabrication processes.

      Free software is not a panacea for all processes in the world. There are very few, if any, solutions that are universal. On the other hand, I may eat my words later as problems have a way of being solved in unexpected ways. There are no intractable problems in the points I've listed, but there may be better solutions to the overal problem than open source.

      Finally, a quick disclaimer. I work for a company that does MIPS-based chip design. So take everything I say with a grain of salt. :)

Slashdot Top Deals

If you want to put yourself on the map, publish your own map.

Close