Cores ported to Chameleon V2

Caution: Non registered users only see threads and messages in the currently selected language, which is determined by their browser Please create an account and log in to see all content by default. This is a limitation of the forum software.
  • So far the other two cores Peter has created have been ported to v2, find them here:


    http://wiki.icomp.de/wiki/Chameleon_VIC20

    http://wiki.icomp.de/wiki/Chameleon_2600


    Also ported are the examples in the public github repository


    Convey's Game of Life

    Hardware test core


    And the following 3rd party cores:


    Chameleon Pong by Alastair M. Robinson

    Minimig ported by Alastair M. Robinson (Thread)

    PC Engine ported by Alastair M. Robinson (Thread)

    OneChipMSX ported by Alastair M. Robinson (Thread)

    FPGAgen ported by Alastair M. Robinson (Thread)

    PACE Arcade cores (Pacman, Pengo, Moon Patrol) ported by Alastair M. Robinson

    Atari800XL by Mark Watson

  • Tobias : I thought the cores that ran on V1 also would run on V2 and that you chose the current FPGA chip so that everything from V1 would work seemlessly on the V2 without any porting. But, if I understand the post above correctly the core authors (e.g. minimig core) now needs to port it to the FPGA used in the V2 ? If this is the case, why did you not upgrade the FPGA to a bigger one so it would have room for larger cores ?

    This is not critique -- I am just curious :-)

  • Simply said, the "new" FPGA isnt actually new, it's basically the same thing in a more modern package and with changed pinout. We switched because the old one became too expensive. And a bigger one would also be more expensive - and require a lot more effort for porting a core, which right now is very easy (it took me a few hours to port one of the examples the other day, and that was mostly because my knowledge in that area is close to zero :))

  • Right now i am porting the examples, and depending on how hard it turns out i will either contact the authors of the other cores or just port them myself :)

  • I'm in contact with more core authors - those who have published cores, but didn't port to the Chameleon yet. I will only be able to support them with the V2 hardware, but since our promise to existing Chameleon customers was that we'll continue to support the old hardware, it's a good idea to keep the FPGAs "mostly the same". So when we get a new core for V2, it'll most likely be easy to back-port it to V1 and make it available to existing customers without too much hassle.


    "larger FPGA" alone won't help making more cores available. It's really a system design that you need to do - interfaces, memory and bandwidth is what you need to look at. Imagine the Chameleon running a core that allows running a flavour of Linux, and 32MBytes quickly become a tight limit. Imagine the VGA output displaying 1280x1024 pixels hi-colour, and you're already using more than two thirds of the memory bandwidth of said 32MBytes. These are just small examples of what to look at when you're designing a new platform.


    Chameleon is a "tried and true" platform. Here at iComp, we don't like to drop support for things that are still great for being used in everyday life. After all, "Retro" is what we do :-)

  • They will get created when you build the core on the command line, like:


    Code
    1. ~/altera/18.1/quartus/bin/quartus_sh -t chameleon2_life.tcl


    perhaps we should check them in, i also fell for it :)


    BTW, i am almost ready with porting your pong core (that was yours, right?) - send me a pm with your email and i'll mail it over... some details dont work yet, but thats related to my complete lack of vhdl/quartus knowledge =)

  • They will get created when you build the core on the command line, like:


    Code
    1. ~/altera/18.1/quartus/bin/quartus_sh -t chameleon2_life.tcl


    perhaps we should check them in, i also fell for it

    Oh OK, yes that works. It might not be a bad idea to include them, because I can't see any way to run that script from the GUI, and running it from the command line is a bit of a pain when you need multiple versions of Quartus on your system, so can't have it in your path.

  • Too bad about the Spectrum..I spoke to the author and whilst he didn't promise he'd work on it he implied that if no one else updated it he would. Oh well, good news that the other person has one as an alternative.


    Fair play about the ZX81 and C128. But I think the ZX81 might be a good goal for myself to try to learn how to this. I write software for a living but hardware is mostly magic to me so it's probably not realistic but i might try.


    After the ZX Spectrum, what I really truly want is the }{gs and the ST. And a finished Vic20 ;)


    What level of CPU is a realistic goal for this? Could it do a 68020 or 030?


    I'd love a Sun3 more than anything else.

  • The Spectrum core for the V1 hardware is kinda difficult to port to V2, as it does not use our Chameleon IO instance, but "something self-made", which would have to be ported to the new hardware. I have my doubts that anyone else than the original author can do that in acceptable time, but the original author has no interest in that core any more

    Is the source for that core publicly available? In a moment of curiosity I could only find binaries.


    Since there are now many, many FPGA cores out there, it might be a good idea to maintain a list of known cores and their current status, so anyone interested in tackling a core can see which ones are in progress and which ones need a port?

  • Quote

    Is the source for that core publicly available?

    Unfortunately not. The author does not likey to publish it (which we respect, of course). I have the source here, but i am unable to do this port. If you are really interested in doing this, perhaps you can contact the author and ask for it? :)