MiniMig with AGA

[Piru]

The 64 MB of DDR SDRAM is configured as 2 MB Chip, 1.5 MB Slow, 8 MB Fast and 48 MB of extra memory which can be used as additional Chip memory. You can have 50 MB of Chip RAM in total.

I guess that chipmem size modifier works by adjusting mh_Upper and calling FreeMem() for the area to add? (that's how I would do it)

I have to say this is quite nice stuff. Now it just needs the 020 core.

[Piru]

The 020 has no data cache. Also the 030 supports a burst mode which the 020 doesn't. Should increase the throughput a lot. But I don't know how these things behave in a fpga reimplementation.

I doubt softcores are going to implement any cache anyway. They just rely on the brute execution power, faster memory etc.

[Piru]

The Natami

Can we please keep to something that is going to happen during this decade (or at all)?

[Piru]

Yeah cuz FPGA tech will never be able to run a Amiga software...

Excuse me? Who claimed anything like that?

[Piru]

Everytime *Natami* comes up, you have something negative to say.

I've questioned the feasibility of the Natami project. I don't believe they will ever finish it (with the plans they have). It's just way way too much work. Just the software side of the project is enormous (10 years or so with moderate sized team).

It does not equal to "FPGA tech will never be able to run a Amiga software" however, as MiniMig has clearly shown. MiniMig project uses the correct approach of dividing the problem to smaller sections which can be solved individually.

Now that we have Minimig and FPGArcade, is it such a stretch to see a similar device with even more oomph?

Yes. I've skimmed the reasons above, but I can dig up the more detailed posts about it if it is required.

Recall the flak Dennis took on this very forum when he first announced his project.

I had my doubts, but didn't voice them. That was because it was clearly within reason to create such a feat. With Natami it is just totally unrealistic. Unless if they cut out most of the unrealistic stuff it will never finish.

As you are (iirc) a MorphOS developer, my feelings are that such a project succeeding may cause a conflict of interest for you here.

You're correct that I'm a MorphOS developer.

You are incorrect in that my dislike of Natami would have something to do with MorphOS. Natami wouldn't compete with MorphOS in any way as they're targeting different audiences. Getting either one doesn't exclude getting the other.

Not every move I make or thing I say have some hidden MorphOS agenda behind them.

However, you could put your talent to use and join the team and write new libraries to take advantage of the platform.

I was already asked years ago. I declined as soon as I realized what kind of unrealistic mess the project was.

When I was contacted regarding MorphOS project ~10 years ago I instantly knew the people behind it were serious and had it all well thought out. I knew it would work and that's why I joined up.

[Piru]

To use the massive throughput of an SDRAM, you have to put a cache.
Pure random access on an SDRAM is not better than an old DRAM : 60ns at best.

Forgot about that. For some reason I built an analogy in my mind that this would be similar to running software emulation on top of a host system, but indeed that is not the case. Here we have to deal with the memory interface directly, which indeed is quite slow. Just shows how much of a software guy I really am.

I have a TG68 softcore with a 4-way 8KB unified cache running on an FPGA.
What's wrong with that ?

Nothing.

So I assume the 68000 softcore has some kind of built-in cache to compensate for the slow memory interface, too?

[Piru]

I think you both need to get to market, if only to see the look on Piru's face.

We'll see what happens...

[edit: thanks kolla]

[Piru]

BTW, do you think it would be possible to emulate the registers of some gfx chip like Virge/Trio64/CirrusGD54xx? If an amiga gfx card was emulated no drivers would be required so we would have P96/CGX support. That would require that 020 emulation was already present.

That shouldn't be too hard, I believe. Those old chipsets are quite well documented, too. The fun part is that (as far as I can tell) it wouldn't require any changes to the HW side of things.

Being able to use existing drivers would be a big plus, indeed.