Porting Mega65 to MiSTer?

[manni199]

Are members here interested in having a Mega65 core on the MiSTer platform?
Is somebody interested in porting the Mega65 to the MiSTer platform?
Alexey Melnikov (sorgelig)?

For those who don‘t know what Mega65 is, look here www.Mega65.org

The successor of the C64 was in development in 1989/90, but never finished.
The Mega65 project is doing exactly that. They took C65 prototypes, analysed them, and finished the C65 on an FPGA system.
Result is the Mega65, that shall be sold maybe next year to interested people.

100 prototypes/developer systems are sold already and will be shipped until end of this year.
https://shop.trenz-electronic.de/en/303 ... t-computer

Maybe it is an idea to port the Mega65 to the MiSTet platform?
This would bring a bigger user/developer base to the Mega65.

The main Mega65 core developer likes the idea that Mega65 gets ported to the MiSTer platform.
He would give full support for such a project.

It‘s far beyond my sklills to do that on my own.
It‘s only that I‘d like the idea to have a Mega65 core on the MiSTer platform.

BTW, the Mega65 is not a cheap system. Estimations go to >750€ or more per system.


Have a look here:
https://www.forum64.de/index.php?thread ... ost1582694

[tontonkaloun]

That would be great.
I wanted to buy this mega65, but found it expensive. I look forward to the public version.

[Alkadian]

@ manni199 Uah! Such a great news! Thanks for sharing!
I just cannot wait for it. Big fan of c64 here and I am very excited about the Mega65 core.

[manni199]

Until now this is only an idea, nothing more.
An I don‘t have the skills to port the Mega65 core to the MiSTer platform.

Until now I have found out that the Mega65 main core developer (Paul Gardner-Stephen) also likes the idea.
He is willing to give support for such a project.
But he has no time to do this port on his own.

And I want to find out, are there people here interested to have this Mega65 core on the MiSTer platform.

Until now many people read my post, but only two answers until now.

And the Mega65 core is not finalized until now.
So before starting a port, there should be the final Mega65.
Some day next year....

The Mega65 project is already on GitHub (github.com).
Sourcecode, Documentation, etc. is available there.

And every time when I have a look at the already existing Mega65 documentation I am deeply impressed.
Download the five PDF files and have a look:
https://github.com/MEGA65/mega65-user-guide/releases

manni199

[NegSol]

Not sure if you are aware of it but the topic was discussed earlier even back in 2017.
https://www.atari-forum.com/viewtopic.p ... 65#p331658

Also Sorgelig did once try to port it already and commented on it. Not sure if the source of the core changed a lot since then.

The quasi nonexistent software library holds down the excitement of most users and devs I would guess.

[manni199]

No, I didn‘t know that.

I started with MiSTer two or three months ago.
Before I read about the Mega65 project and the extreme successful ‚Spectrum Next‘ kickstarter campaign.
There are many people interested in the Mega65.
The estimated price for the Mega65 is quite high (750€ or more).
So I thought having the Mega65 on the MiSTer platform would be a good idea.

Until now the Mega65 is not finalized.
So best is to wait for the finalized version.
Maybe then there are more people interested.

The complete Mega65 project can be found at github.com

[Darthcustom]

I really support having Mega65 core ported to MiSTer.... I'm following Mega65 for long time but 750€ is too high to my taste. DevKit was 1000€. I´d see it more in the 500€ price range.

I'd love to see Mega65 running in my MiSTer....

[kathleen]

Same here, the Mega65 ported to the Mister will be a dream, the "real" Mega65 being far out of my budget, getting it on the Mister would be more than nice

[Sorgelig]

Let's wait for final version first.
Unfortunately C65 has no software. Even ROMs weren't final. The only thing in C65 was useful is C64 mode. So the only command you need in C65 is GO64.
As for C64: MiSTer already has very compatible C64 core and i don't expect the one in Mega65 beat it in any aspect.
Same as with Spectrum Next, it's useful as a whole device. Mega65 has nice case with keyboard (i wish they will sell cases separately!) and boot ROM managing specific HW. MiSTer doesn't need all these addons as MiSTer has it's own API servicing C64 perfectly.
While port of Mega65 is possible, it won't bring anything new.

So there are 2 main things before port:
1) wait for final version
2) check the requirements because later versions start to use quite large FPGA with a lot of BRAM - much more than DE10-nano has.

[manni199]

great, I was looking forward to your statement about porting the Mega65.
You are the administrator here, so I was almost sure that you already follow this thread.

One question to you Alexey, how many manhours/mandays would you estimate it takes to do this port?
I am sure you already had a look on the HDL sources on GitHub?
Are there very big differences between MiSTer and Mega65 platform, so that there ist much more to do than simply recompile the Mega65 sources?

I completely agree with you that there must be a finalized Mega65 before starting a port.
Especially because the Mega65 is an enhanced version of the C65. So all new features must be finalized.

I thougt RAM would be no problem, because at least 32MB SDRAM is mandatory fo MiSTer to use all cores?
An the ARM core with 1GB RAM (Linux system) on the MiSTer board should compensate any disadvantage to the Mega65 FPGA.
It’s because I have no skills in programming HDL/VHDL you should judge the technical details to compare MiSTer with Mega65.

Also very important in my eyes is, that Paul Gardener will give any support that is needed to port Mega65 to Mister.
And Paul will applaud to have a Mega65 core on MiSTer.
I know that man people want a Mega65, but when the hear the estimated selling price (>=750€) the interest cools down.

To say it again, I would like having a Mega65 core on MiSTer because ist generates (hopefully) a much bigger user base for Mega65.
More user will result in more software....

Also every time I have a look on the Mega65 documentation I am impressed.

And the MiSTer platform is much much cheaper than Mega65. Mega65 is more a sculpture to put in a showcase to enjoy looking at it.


I know, the Mega65 designer want us to use and work with it for sure.
But I‘m sure there will be people that use MiSTer for use/work, and put the Mega65 in a show case (once when there is a Mega65 core for MiSTer).

So let‘s wait for mid of next year when final Mega65 is released.
...

[Newsdee]

To say it again, I would like having a Mega65 core on MiSTer because ist generates (hopefully) a much bigger user base for Mega65.

If they need that badly to grow they user base, they could port the core themselves, no?

[LamerDeluxe]

I'd love to see this ported to the MiSTer, if at all possible. I've been following the C65 with interest since the first news about it.

[limi]

Mega65 has nice case with keyboard (i wish they will sell cases separately!)

They will! More details here.

[dave18]

Core still has non synthesisable code which Quartus will refuse to compile. The Cpu core might need a rewrite to make it portable and as it is a custom Cpu there is no drop in replacement. Didn't look at whether the largest chunk of bram could be moved to sdram as it is pointless while the core has code that can't be synthesised.

[dave18]

Ignoring the fact that there is non synthesisable code, I think the Mega65 design principle is a problem, it seems like they've just chosen a high end FPGA and taken the easy option of using the on board memory for all the different RAM/ROM needs. It makes the core creation easier as they don't have to worry anywhere near as much about contention and bus sharing but the downside is

1) Makes the product much more expensive - so little take up of the actual hardware
2) Makes it difficult to port to other hardware as not many other systems will have an FPGA big enough and the effort to port all the FPGA RAM to SDRAM/DDR etc is difficult and unlikely to be completely compatible.

It just feels like the Mega65 will never gain enough users for any meaningful development to take place and that seems to me that they've got the strategy wrong.

I think the Mega65 is probably not going to make it to MiSTER as the effort/reward doesn't stack up but be glad to be proved incorrect.

[ericgus09]

It just feels like the Mega65 will never gain enough users for any meaningful development to take place and that seems to me that they've got the strategy wrong.

I feel the same way , sure you will get some interesting demos and maybe a game or two .. but thats it .. it likely will be one of those interesting hobby fantasy machines (like the CommanderX16 etc) where I would poke at it for a while then put it up on the shelf .. if I were to use it, as was mentioned by someone a while ago would be typing "go64" .. and for that the Mister is far better..

That said.. what DOES interest me with the Mega65 is the physical housing/casing/keyboard .. that is of interest to me, not so much the actual system board. There was never a real commercial release of the c65 .. so there is no software base.. and the mega is just someones interpretation of that machine filling in the missing gaps of an incomplete product so in that its a partial fantasy system..

[Darthcustom]

At the right price, the Mega65 can be worth it even if it is just a C64 with integrated SD slot, disk and 40 Mhz turbo.

[DoubleA]

At the right price, the Mega65 can be worth it even if it is just a C64 with integrated SD slot, disk and 40 Mhz turbo.

Well "right" is relative and subjective. But "high" seems to be set already .

[Moondandy]

Hey guys, thought I would necromance this thread now it seems the C65 is finalised and coming out fairly soon. Does this core look any more portable now? Is anyone in touch with people on the C65 team who would be interested in porting it to MiSTer?

It looks to have the potential to be an interesting system if it takes off, but the actual hardware isn't cheap. MiSTer core would be nice, if doable.

[akeley]

~800 Euros if you want one in Europe, oh my

[R4M]

You could also get a digilent Nexys A7 evaluation board for around 250 Euros. It has a smaller FPGA (100T vs. 200T), but the team still maintains a current Mega65 core for it. They also contemplated selling the MEGA65 board separately after the first batches of the complete computer are sold.

One reason I won't buy a Mega65 is its insufficient amount of RAM. They only have ~9 MB (1MB block inside FPGA, an 8 on an external chip), which is not really sufficient for 16/32 bit cores like Amiga/Minimig.

[Blitzwing]

I thougt RAM would be no problem, because at least 32MB SDRAM is mandatory fo MiSTer to use all cores?
An the ARM core with 1GB RAM (Linux system) on the MiSTer board should compensate any disadvantage to the Mega65 FPGA.
It’s because I have no skills in programming HDL/VHDL you should judge the technical details to compare MiSTer with Mega65.

Sorg was talking about BRAM (block RAM) it is the embedded memory inside the FPGA itself, it’s separate to the external DDR and SDRAM. The Cyclone V in the De10-nano doesn’t have much.

[Swainy]

I see that some people have started receiving their Mega65’s. I assume that the software is final now?

[riker77]

I see that some people have started receiving their Mega65’s. I assume that the software is final now?

Hi,

the MEGA65 is a project for the decades. You might say, that the hardware is more or less final, but the software (FPGA-Core and ROM) is probably years away from being finished. This is and probably will never be a consumer product like your minified emulator boxes or the Analogue devices. I like what I got, because I knew what I was getting into, but the maturity isn't there yet.

[Swainy]

Well I guess that the same could be said of the ZX Spectrum Next, which I own but hardly use since the core was released on the MiSTer.

[pepito]

MEGA65 is a FPGA machine, so the core should be very easy to port out to Mister.
someone have to buy one

[ByteMavericks]

MEGA65 is a FPGA machine, so the core should be very easy to port out to Mister.
someone have to buy one

Well volunteered! Good luck with the code that’s already identified as non synthesisable

[cptmadness]

MEGA65 is a FPGA machine, so the core should be very easy to port out to Mister.
someone have to buy one

Well volunteered! Good luck with the code that’s already identified as non synthesisable

You mean the code that was identified as non-synthesizable by Sorgelig in 2017? The same code where he himself said "I think it still needs couple years to wait for well organized code. Then it will be possible to port to other boards."

Quick everybody, stop trying. ByteMavericks says it can't be done.

[ByteMavericks]

Do it, prove me/Sorgelig/everyone wrong! You’ve claimed it’s “very easy”. I’ll be the first to buy you a beer/beverage of your choice.

[cptmadness]

I made no such claim. The only thing I’m saying is that the information you’re going off is 5 years out of date, and sorg never said it couldn’t be done, just that it couldn’t be done at the time. As has happened time and time again, almost everything claimed to be “impossible” for MiSTer has appeared after some amount of time. I just find the constant naysaying to be draining and a weird and deeply entrenched negative aspect of the MiSTer community that has been shown to be ridiculous on many occasions, and yet it somehow persists.

You will be proven wrong on this at some time by someone, of that you can be absolutely certain.