Welcome, Guest. Please login or register.

Author Topic: NatAmi 68070 design draft  (Read 36873 times)

Description:

0 Members and 1 Guest are viewing this topic.

Offline CodeSmith

  • Sr. Member
  • ****
  • Join Date: Sep 2002
  • Posts: 499
    • Show all replies
Re: NatAmi 68070 design draft
« on: June 26, 2008, 08:27:42 PM »
@bloodline

Quote
No, I separate out the NATAMI because unlike the MiniMIG and the CloneA, the motivation behind it don't make any sense, or at least seem very unrealistic!

The motivation is making something cool and selling it to a few like-minded people who won't be able to build one themselves.  I don't understand why you think there's a difference between the motivation behind the MiniMig and that behind the NatAmi.  The CloneA *is* different - Jens wants to basically clone the old Amiga chips and sell them by the million.  Dennis and Thomas are doing it for fun.  Don't be misled by them wanting to make a few bucks on the side, I never got the impression from them that they're out to do anything other than inject a bit of life back into the Amiga hobby.

Quote
Every extra feature takes time and silicon away from debugging the standard AGA features... Plus somethings I have read about the "Super" features would be incompatible by design... anyway that isn't the point, why bother with these new features, no existing software can use them!

According to Gunnar, the "super" features have been designed to be backwards compatible.  Presumably they did this the same way you can add A1200 style clock ports in a compatible manner to a C64, ie by using memory addresses tagged as "future expansion" in the hardware docs.  Since CBM's no more, it's safe to use those.  I agree that no existing software makes use of the new features, but again I point you at the C64.  There's a very cool game called "metal dust", that won't run on a stock C64 - it requires a SuperCPU expansion that was released after CBM went under. The SuperCPU is an accelerator that adds a new, partially incompatible CPU (undocumented opcodes don't work and it's not cycle exact) to a C64, so it's a significant hardware change.  People will write new software for an old platform.

Quote
I don't care how well it works, I just don't understand the project motivations or perhaps really I don't understand what benefits this has over MiniMIG and CloneA as a commercial product... but I can see disadvantages...

You see the NatAmi as a commercial project that follows the normal rules of business, and it that sense it's insane.  But then again, the MiniMig is also insane from a commercial point of view.  How many have been sold so far?  I'd be very surprised if the total number of units "out there" breaks three digits.  Think of the NatAmi as the A4000 to the MiniMig's A500 and it will make a lot more sense.

I can understand your confusion though, there are a few posters here, at aw.net and at natami.net who are making some pretty outrageous claims about what the hardware can and can't do.  The NatAmi team consists of two people right now: Thomas and Gunnar, and their claims are a lot more scaled back.  Anything said by anyone else is probably suspect.

 

Offline CodeSmith

  • Sr. Member
  • ****
  • Join Date: Sep 2002
  • Posts: 499
    • Show all replies
Re: NatAmi 68070 design draft
« Reply #1 on: June 28, 2008, 06:38:16 AM »
@Hans_

The NatAmi team described their roadmap on their forum, and the new CPU is supposed to come in only after the first version has been released.  I think the recent chatter around the CPU arose from their deciding against coldfire and posting some ideas, and the discussion that came about around that (I can't find a link, but Gunnar mentioned that right now the CPU is at the stage where he and Thomas are just kicking ideas around at lunchtime; the main event is still the '060 board).  Gunnar has explicitly said that the 68060 version will come first, and the CPU will follow only once the SuperAGA chipset has been debugged.  Hopefully the first release will happen sometime in the next 2-3 months (summer 2008), but given the nature of engineering projects I'm guessing it will instead happen in the last 2-3 months of the year.