Hallo Besucher, der Thread wurde 6,6k mal aufgerufen und enthält 30 Antworten

letzter Beitrag von Snoopy am

Porting the Mega65 project to MiSTer platform?

  • I‘d like the idea that the Mega65 Project gets ported to the MiSTer Project.

    That would increase the number of Mega65 users by far.

    It seems there is a very big MiSTer Community.

    And it‘ll make the Mega65 much more affordable.

    I was told in this forum that the final Mega65 will be quite expensive.

    Very high quality and very well designed, but expensive.

    (750€?)


    A base MiSTer costs <200€.

    Prebuilt MiSTer can be bought in several internet shops.


    I like the idea very much.

    Amiga 500, Amiga 1200, Acorn Archimedes, Atari ST/STE, PC 486 and hundred systems more (computer/consoles/arcade) are directly aviable.


    igel65

  • It is questionable if the FPGA of MiSTer is powerful enough to port the MEGA65 there. There was a question once whether a C128 can run on a MiSTer. That was a similar problem.

    ___________________________________________________________________________
    Ultimate64, TAPunio, SD2IEC, ZX Spectrum 48k, 1581 Replik, C64 Laptop, C64 MK II, C116, SX64,
    MiSTer FPGA, TI99/4A mit PEB, Atari 800 XL, Anycubic I3 Mega, Mega65, C64 Modular, Uniprom64

  • It is questionable if the FPGA of MiSTer is powerful enough to port the MEGA65 there. There was a question once whether a C128 can run on a MiSTer. That was a similar problem.

    This I don't believe at all. Maybe the Cyclone III in the original MiST is not enough, but the Cyclone V SoC in DE10-nano is quite powerful. I'm too lazy to look up the specs right now and compare to the Artix board that is used for MEGA65 development, but just think of the machines, MiSTer can easily recreate: AA-Amiga, ao486 PC, SNES, etc.


    And a C128 shouldn't be to complex. It's just an slightliy extended and faster 65xx and a Z80, VIC, SID and VDC. Nothing to complex. I guess even a MiST has enough LEs to stuff a C64 and CPC or Spectrum core together and this would be more complex than a C128. I guess the only reason there is no C128 core for MiST or the Mister is, that there is no real demand. Most people used the C128 with GO64 first.

    Most Retro-FPGA users are mostly interested in games and there are no killer games for C128 only. And for the CP/M mode and Z80, there are enough alternatives to C128.


    So if I don't miss something important, a C128 core should be doable on inexpensive FPGAs. It's just that no one did it yet.

  • I‘d like the idea that the Mega65 Project gets ported to the MiSTer Project.

    They are also talking here about this idea: misterfpga.org

  • The 200T is four times more powerful, than the FPGA in the MiSTer.

    ___________________________________________________________________________
    Ultimate64, TAPunio, SD2IEC, ZX Spectrum 48k, 1581 Replik, C64 Laptop, C64 MK II, C116, SX64,
    MiSTer FPGA, TI99/4A mit PEB, Atari 800 XL, Anycubic I3 Mega, Mega65, C64 Modular, Uniprom64

  • Zitat von wins4

    The 200T is four times more powerful, than the FPGA in the MiSTer.



    OK, too much guessing. I always am confused which Artix A7 board is big enough for MEGA65, but from what I found on this forum, already the Artix-7 FPGA XC7A100T should be fine.

    So let's compare it with the Cyclone V SE from DE10-Nano/Mister:


    Artix7 XC7A100TCyclone V SE 5CSEBA6U23I7
    #LABs/Slices1585013750
    CLB/ALM-Config4x6-Input, 8 Flipflops2x4 or 1x 6,7,8-Input, 4 Flipflops
    approx #LUT495100110000
    max Operating MHz450 MHz800 MHz
    BRAM4860 Kbit5570 Kbit
    #MULs/DSPs240224


    Does not look to different, also it is always hard to compare different layouts/modes of logic elements. This comparison to 4-bit lookup tables is always a bit misleading and surely influenced by marketing departments of Xilinx/Intel.


    The MEGA64 dev kit seems to have taken the 200T variant, but I don't think that it is really needed for the creation of any 8 bit system. Even complex 32 Bit systems like Minimig-AGA/RTG with fast 68020 32Bit CPU fit into the 25000 LUT4 LEs of the Cyclone III in MiST. The TG68 runs with ~100MHz in the Mister and performance rivals a A4000/EC030. I can't believe that a MEGA65 is more demanding.

    At least for the C65 we are talking of a 8Bit low-cost system Commodore planned to rival NES/MasterSystem for people who couldn't afford an Amiga 500 in 1989/90.

    Not sure if the MEGA65 mode extensions make it to rival 8/16 Bit SNES or16/32 Bit OCS-Amiga, but surely not a A3000/4000 with ECS/AGA/RTG and 25MHz 68030 32-Bit CPU.

  • RexRetro

    one question.


    The MiSTer FPGA (Cyclone V SE) does contain a dual core ARM CPU (900MHz, 1GB RAM) that makes this FPGA very flexible and useful for emulation/simulation.

    Every IO operation can be handled over USB with a ‚linux driver‘. So connecting controller, keyboard, etc. can be done very easy.

    And parts of the 1GB RAM can be used for the emulated systems.


    Does the Artix7 XC7A100T (or 200T) contain also an ARM CPU?

  • RexRetro

    one question.

    Does the Artix7 XC7A100T (or 200T) contain also an ARM CPU?

    No. Maybe this is a reason to go for a bigger FPGA in the MEGA65 dev kit, so it has enough space to fit an additional "management softcore" (maybe something like RISC-V, that has modern compiler support and can run Linux).

    But there are also FPGA/ARM-hardcore SoC from Xilinx, like the Zynq-7000 series. There are inexpensive ZYBO boards similar to DE10-Nano from Digilent.

    ZYBO comes with dual Cortex-A9 @ 866MHz, 512MB DDR3, HDMI+VGA, OTG USB2, Gigabit Ethernet, Audio Codec, MicroSD, many Pmod connectors.

    But with 28000 LEs, 2.1Mbit BRAM and 80 DSPs the FPGA part is more in the area of the Cyclone3/4 used in MiST/SiDi, and I guess it would be more complicated to use (neighboring) PMods as GPIO for extra SDRAM.

  • RexRetro


    thanks for the explanation.


    Hopefully the Mega65 project will be ported sometime to the MiSTer project.

    It would raise the Mega65 community by far.

    The MiSTer hardware should be capable of it easily.


    I‘d like a statement from a main Mega65 developer about it.

    (Paul Gardner-Stephen?)


    Maybe the developer of the MiSTer C64 core would love to port the Mega65!
    Is it an idea to contact them?

  • Hi Paul.


    Great that you are interested in porting the Mega65 to the MiSTer Platform.

    (supporting members of the MiSTer team to do this port).


    I just had a look who are the members on GitHub for the MiSTer project.

    It seems that Alexey Melnikov (sorgelig) is working for the C64 core on MiSTer.

    Maybe its an idea that you contact him directly.


    Would you be so kind and make a little statement here about the reaction of the MiSTer project members for this port?
    No matter if they are interested or not.


    Maybe I‘ll start a thread in the MiSTer section here on ‚forum64.de‘ to get some reactions from other members. Only to get an idea how big the interest is for porting the Mega65.


    I whish I could do the port, but I have no experience at all in programming HDL/VHDL.

    I am a programmer, but for industry PLC controller (Siemens S5/S7, CoDeSys and some special SCADA systems). On Windows/Linux/Mac OS I am more ore less an experienced user. So nothing really helpful.

    Sorry


    BTW, I am from Germany.

    So sorry for my english skills.

  • Hallo,


    kein Problem, dass Englisch nicht deine Muttersprache ist. Ist sicher auch bemerkbar, dass Deutsch nicht meine ist ;)


    Ich wäre ganz froh sein, wenn jemand der MEGA65 an den MISTer portierte, aber leider kann ich nicht der Jemand sein. Ich bin schon total beschäftigt mit der Öffentlichen und Fertigung des MEGA65s. Aber ich würde was ich leisten könnte, an Hand jemand, wenn man Lust und Zeit hätte, daran zu arbeiten.


    LG

    Paul.

  • Hallo,


    Ich kann Alexey Melnikov mal anschreiben (versuchen).

    Das Ergebnis werde ich hier posten.


    Letztendlich wird sich aber jemand vom Mega65 Projekt mit jemanden vom MiSTer Projekt kurz schliessen müssen. Mir fehlt dazu jede Fachkompetenz.


    Eine Frage an dieser Stelle.

    Der C65 existiert ja nur als Prototyp, wurde also nie fertig entwickelt.

    Habt ihr denn alle Features des Mega65 festgelegt und auch schon implementiert?

    Ist der FPGA core des Mega65 final fertig?


    Eine Portierung zum MiSTer macht erst Sinn wenn die Entwicklung des Mega65 abgeschlossen ist.


    Aber es ist sicher eine gute Idee, wenn sich Mega65 Entwickler mit Leuten vom MiSTer Projekt kurz schliessen.

    Ich kann das Ganze auch nicht wirklich beurteilen.


    igel65

  • The MEGA65 core is "mostly" finished, but there are still bits and pieces being worked on. For porting to the MISTer board, the main outstanding item is a rework of the memory sub-system, that I am hoping will be done in the next few weeks. This should make it much easier to port to different boards that don't use Xilinx parts, or use external SRAM etc.


    LG

    Paul.

  • Hi Paul.


    I just had a look who are the members on GitHub for the MiSTer project.

    It seems that Alexey Melnikov (sorgelig) is working for the C64 core on MiSTer.

    Sorgelig started the Mister project and is the main guy for most Mister cores.

    But he ported most from MiST, what should be not too much effort, as he planned Mister as an improved MiST, using a newer Cyclone FPGA and base all on extending a stock Intel starter board instead of a custom designed board like MiST.

    Big missing piece was the 32MB SDRAm that was added via GPIO pins. Big selling point is HDMI instead of VGA with different syncs for different cores.