MiniMig with AGA

[freqmax]

I hope the effort is concentrated on getting the boards free of bugs out the door

Core:s, especially 8-bit ones shouldn't even be a challenge.

Maybe the visual6502 projekt takes a look at the m68k die. It certainly could be nice to have decoded.

[little]

Since the spartan 3e wont give use speeds beyond 20Mhz (or so I read in a prior post) wouldn't it make sense that someone worked in porting the DCTV hardware (that would allow high color resolutions without requiring lots of memory or bandwidth) and hopefully someone would write a RTG driver for it. The Phillips CDi used such screen modes for its gaming library, who knows, maybe some programmers will find it attractive to write ports of games to such a low spec system (but with high color output). That bundled with the AHI sound system should be a nice combo IMO.

[gaula92]

We also have C64, spectrum, zx81, apple 1, cray 1(!) and lots of home games consoles being worked on.

THIS Cray-1 --> http://en.wikipedia.org/wiki/Cray-1

This thing looks like it came out of TRON!! WOW!!
I'll be careful not to work with it while it has some digitaliser attached, I don't like hi-speed motorcycles...  
I can't wait to put my hands on a thing like this, running on my room in FPGA!

[yaqube]

Since the spartan 3e wont give use speeds beyond 20Mhz

Actually there is a hope it will work at 28 MHz. I have it currently running at that speed and with a very small cache (for code only) of 16 bytes it outperforms an 030@25MHz.



I'm working on the implementation of a more sophisticated cache subsystem. I expect the performance to be improved a little bit.

wouldn't it make sense that someone worked in porting the DCTV hardware (that would allow high color resolutions without requiring lots of memory or bandwidth)

Sometime ago I wrote a software decoder for DCTV encoded images just to find out how it worked. Initially I wanted to implement this decoder into the Minimig AGA but I think it's not worth the hassle. Let's better spend that time on a real RTG frame buffer implementation.

and hopefully someone would write a RTG driver for it.

I can bet nobody will.

[ferrellsl]

@little

No need for that.  Just modify/extend the existing Minimig core to add the additional resolutions and colors.  That's how AGA was added to what was essentially an A500 softcore. That's the beauty of a softcore on the Minimig (or FPGAArcade).  Now that AGA and additional chipram are available, it won't be long before some enterprsing coder figures out how to add Super-AGA to the same core....and maybe 030 or higher CPUs.  It may require a larger FPGA, but it's possible.

[little]

Actually there is a hope it will work at 28 MHz. I have it currently running at that speed and with a very small cache (for code only) of 16 bytes it outperforms an 030@25MHz.

That is great news!

Let's better spend that time on a real RTG frame buffer implementation.

+

@little
Now that AGA and additional chipram are available, it won't be long before some enterprsing coder figures out how to add Super-AGA to the same core

On the surface it sounds cool to create a new framebuffer, AGA+, etc. but the problem I see with that approach is that by creating new hardware, hardware that effectively only a few have access and that no one has ever worked with before, you reduce exponentially the number of people  that can write code for it. It might sound easy enough to say "We will simply expand UAE to emulate it also", the problem being that emulation of low level features is either imperfect or way too complicated/slow. The only solution I can think for such dilemma would be that the AROS 68k port implements all the required low level tinkering so that there is no need to make calls directly to the hardware, so a program written for x86 aros would work in 68k aros using the new audio/video hardware, totally transparent to the original programmer.

[yaqube]

Results for 28 MHz clock and 256-byte instruction cache. Maybe the speed gain is not impressive but I have yet another option to try.

[RMK305]

Any speed gain is impressive, no matter how small.

[kolla]

Very cool - the speed is comparable with the ACA-630 and ACA-1230/28 - woohoo
I can't help thinking about how neat it would be to see AROS/m68k in its current "infant" state try to boot on this

[ferrellsl]

@little

No one is creating new hardware by extending the Minimig cores.  And backward compatibility is being maintained with previous real-hardware Amigas by both the Natami team and MikeJ on his FPGAArcade board.  It sounds like you may be a bit unclear as to what an FPGA really accomplishes.  The A500 core on a Minimig is simply a complete A500 recreated in software and loaded onto the FPGA chip.  New "virtual" hardware can be created as needed just by modifying the existing core and loading it onto the FPGA. Therefore everyone who owns and uses a Minimig can load the core-system of their choosing, whether it's an Atari, C64, A500, A500 AGA, etc....

And no one is concerned about x86 Aros running 68K Aros programs because x86 AROS was never intended to have binary compatibility with 68K AROS.  x86 AROS uses a variant of UAE to run classic Amiga software and it works just as well, maybe even better than WinUAE since it's integrated almost seamlessly into Wanderer/Zune (the AROS GUI shell).  The AROS devs ARE shooting for 68K AROS and classic-Amiga compatibility, which is well underway with the Kickstart replacement bounty.

And if you find that there's an application that doesn't behave well on one of the modified AGA/SuperAGA Minimig cores, then just revert back to a stock A500 softcore on your Minimig....problem solved.

As for programmers/developers being discouraged and not writing new applications for the FPGA Amigas and their proposed "extensions", well, it may be news to some, but that already happened about 10 years ago.  Viable commercial software development for the Amiga is dead.  Same can be said for next generation "real" Amiga hardware such as the X1000 and the Natami and Minimig.  The Amiga is and will remain a system for hobbyists.  No one is gonna get rich these days on ANYTHING Amiga, least of all programming for it.  But at least with a Minimig, Natami, or FGAArcade people can buy something to relive better times and still have the option to write software for a system that'll be around a lot longer than classic A1200s, A500s, etc.  Most classic Amiga users will admit that their systems are running on their last legs, held together with Scotch tape and a prayer.

[Hattig]

Results for 28 MHz clock and 256-byte instruction cache. Maybe the speed gain is not impressive but I have yet another option to try.

http://www.yaqube.neostrada.pl/images/SysInfo28-256.gif

I think this is very impressive for a core that was merely a 68000 core a short time ago. Is this with the 16-bit data bus still? Would there be any advantage to implementing a data cache?

5x faster than a stock A1200 and faster than an A3000 is not to be sniffed at.

[Linde]

I'm sorry if this has been answered already, but what is missing from m68k/32 for it to be 68020 compatible? I love the speed benchmark results

[freqmax]

To run the majority of the software. The most popular hardware platforms ought to be targeted (A500 and A1200 is my guess).

As for the CPU, maybe something else in the design could be boosted to improve performance.

And FPGA:s don't run any software. Their gates are configured after reset to interact in a specific manner. And imperative languages C++ etc.. translated to VHDL will be a serious compromise.

[little]

@little

No one is creating new hardware by extending the Minimig cores.

Huh? The moment you speak about something better than AGA it is new hardware, plain and simple. Do not confuse the fact that you can program the FPGA to recreate an amiga with the fact that you can add hardware to said computer that never existed before.

And no one is concerned about x86 Aros running 68K Aros programs because x86 AROS was never intended to have binary compatibility with 68K AROS.

I apologize for not saying this before (I thought it was quite obvious but this proves the opposite), I was never talking about x86 binaries running in 68k hardware, I merely was talking about recompiling the source to run them in 68k.

And if you find that there's an application that doesn't behave well on one of the modified AGA/SuperAGA Minimig cores, then just revert back to a stock A500 softcore on your Minimig....problem solved.

Let me spell this clearly, if someone writes a game with 24bit graphics and sound for AROS x86 and manages to recompile it for AGA/ECS/OCS it will look/sound bad. Even if AROS is successfully ported to 68k said game will still look god awful in an improved AGA core if the OS does not know how to access the improved graphics/sound hardware. Asking a programmer to make a game for a hardware platform with a user base below a thousand is sheer madness.

The Amiga is and will remain a system for hobbyists.

No question about it, but hobbyists are the kind of people like the idea of getting some OSS software and recompiling it to run in this supposedly dead platform.

[freqmax]

The question is for whom the platform is dead for

[ferrellsl]

@little

And again,  I say no new hardware is being created at all by modifiying an FPGA core.  FPGA cores are software VHDL code.  There is absolutely NO hardware being created when someone modifies a core.  Please do some more study about FPGAs before you start spreading misinformation.  Creating new hardware would be adding additional ICs, capacitors, resistors, etc.  That does not happen when someone modifies a softcore.  That's why it's called a "soft" core.

No, you are confusing the goals of AROS 68K and classic Amiga.  The goal of AROS 68K is to run classic Amiga software completely unmodified.  The target program will look and operate identically on an Amiga that's running AROS 68K or OS3.X.  In fact, it will be running on identical hardware. Instead of OS 3.X on your Amiga, you simply choose to run Aros 68K instead.  Or if you have an FPGA based Amiga, you can do the same there as well.  Run AROS, or run OS3.X or less.  The software application is still using Amiga hardware, but you can opt for an FPGA Amiga as well.

And I haven't seen nor heard of anyone here asking a programmer to make a 24-bit game for a classic Amiga.....No one here will disagree with you that asking such a thing is madness.  Gunnar is actually making one for the Natami though, now that you mention it.

And who is developing ANY 24-bit screen mode games for classic Amigas anyway? (Answer:  Nobody)  And why would they?  There's simply no market for it, and the CPU horsepower for such a game on a classic Amiga doesn't exist.  So your argument makes no sense.  So the FPGA devs are supposed to keep their next generation Amiga hardware and softcores stunted/handicapped for software that doesn't even exist?  The entire point of next generation hardware is to move forward with greater capabilities while retaining the ability to run older classic software if you choose to do so.