Amiga.org

Amiga computer related discussion => Amiga Hardware Issues and discussion => Topic started by: DamageX on June 30, 2005, 07:39:29 AM

Title: why no updated chipset after AGA?
Post by: DamageX on June 30, 2005, 07:39:29 AM
Who had/has the capability to produce a new Amiga chipset and why didn't it ever happen? The cost of design? of manufacturing?
Title: Re: why no updated chipset after AGA?
Post by: Akiko on June 30, 2005, 07:42:08 AM
I believe the AAA chipset was in development but commodore went bust before completion..
Title: Re: why no updated chipset after AGA?
Post by: bloodline on June 30, 2005, 07:49:20 AM
Quote

DamageX wrote:
Who had/has the capability to produce a new Amiga chipset and why didn't it ever happen? The cost of design? of manufacturing?


Do a google for "Hombre", "AAA" and AGA :-)
Title: Re: why no updated chipset after AGA?
Post by: Linchpin on June 30, 2005, 08:17:38 AM
... and get hold of a copy of the Deathbed Vigil.
Title: Re: why no updated chipset after AGA?
Post by: PaSha on June 30, 2005, 11:10:18 AM
Quote

DamageX wrote:
Who had/has the capability to produce a new Amiga chipset and why didn't it ever happen? The cost of design? of manufacturing?

Why? Ask Mehdi Ali and Bill Sydnes....


-Paul
Title: Re: why no updated chipset after AGA?
Post by: SamuraiCrow on July 01, 2005, 01:57:00 AM
Quote

DamageX wrote:
Who had/has the capability to produce a new Amiga chipset and why didn't it ever happen? The cost of design? of manufacturing?


All of the above.  By the time Amiga was in good stable hands the AAA chipset designs had been lost, they had been written in a non-standard gate-layout language that was proprietary to Commodore, they weren't backward compatible to AGA, they didn't support EDO memory or SDram and were therefore obsolete, and many other reasons besides.
Title: Re: why no updated chipset after AGA?
Post by: billt on July 01, 2005, 03:25:18 AM
Commodore died working on AAA and Hombre which were the next chipsets.

Today, it seems the schematics and other data files have been lost. Presumably Gateway would hold the rights if the whereabouts were known.

As for why not, it's incredibly expensive to do, and with a market our size it's not even remotely possible to recoup such expenses from sales.

To print the mask set in 0.18 micron can cost over US100,000. To design the circuit will take some time and salary. The CAD tools to do the work will cost hundreds of thousands of US$. Expect at least one respin to fix bugs. A *cheap* chip can cost around US$1 million to put into production. How many chips do you want in your custom chispet? Can selling 1000 or so units pay for your chip expenses, let alone the entire system??

Financially, there's no possibility of it.

Technologically, off-the-shelf parts have far far surpassed AGA, and there's no way we can catch up with them to offer a meaningful product. NVidia spent about 1/2 Billion US$ to develop their GeForce 6 series chip. Some people say ATI has trouble keeping up with the GeForce 7 series coming soon, and ATI has a ton of cash. What could the Amiga market possibly do to offer an acceptable competing product like that?

It does not make sense to make new chipset today, as nothing custom could possibly compare well with the off-the-shelf standards.

I myself would prefer to take the best technology possible. As we cannot possibly do better than Nvidia or ATI or Creative Labs or Envy24 or whatever, then we should embrace the best standard parts we can possibly use.