Amiga.org
Amiga computer related discussion => Amiga Hardware Issues and discussion => Topic started by: freqmax on December 17, 2011, 01:21:52 AM
-
Which nm manufacturing process has been used to produce the Agnus/Denise/Paula chips in Amiga 500..?
And what transistor count and die size is it?
(The scanning thread seems to went about something else)
-
I would guess 2000nm to 1500nm for process... And I doubt any Amiga custom chips has more than 30,000 transistors.
-
Jut looked up MOS (later CSG) and I think they were stuck on some pretty old technology... My guess is that the Amiga chips are probably even 3000nm (3micron)...
-
I wonder if a digital camera can get enough detail from a 1.5 µm chip to deduce logic?
-
Jut looked up MOS (later CSG) and I think they were stuck on some pretty old technology... My guess is that the Amiga chips are probably even 3000nm (3micron)...
Wasn't some parts of AGA fabbed by HP? Presumably because CSG couldn't fab what was required.
-
Meet...
(http://www.amiga-resistance.info/display_photos/lisa_small.jpg)
Lisa.
-
Wasn't some parts of AGA fabbed by HP? Presumably because CSG couldn't fab what was required.
I think you are right, CSG's fabs just weren't up to the task, Lisa might even have been some standard HP mixed mode ULA (now there's a thought for another thread!;) )...
I also seem to remember Jay Miner expressing surprise in an interview in the early 90s that most of AGA was still fabed on the original process that they used back in the mid 80s...
-
I wonder what nm process the AGA chipset might have used..
-
I remember reading somewhere that the original 3 magic chips were supposed to be one big chip, but the fab tech of the time was not up to doing it all on one die.
-
If anyone knows the transistor count and die size. It might be possible to do some estimation.
-
Anyone else remember that aga was to be intergrated into the cpu die? I remember a few interviews with key figures (David Pleasance or Lou Egbert perhaps, it was a while ago) in amiga magazines back in the day talking about it and I was curious as to how they'd planned to do it being that they had no license for 680x0, nor did they have the facilities to do it even if they did.
Both forward thinking and stuck in the past at the time (ie. intergration of custom hardware in a cpu die at the time was pretty forward thinking, but sticking with aga was kinda crazy).
-
@fishy_fiz: I think you are confusing AGA with Hombre. AGA was never much more than an ECS upgrade. Hombre was supposed to have a PA-RISC core integrated in the chipset.
-
I wonder what nm process the AGA chipset might have used..
I already pointed out that it would be either 3000nm or 1500nm....
http://en.m.wikipedia.org/wiki/3_µm_process
-
Nah, it was definately aga. I remember arguing with my brother about the advantages it would provide. He was of the opinion it was interesting, albiet a little silly, and I was such a hardcore amiga guy, and too young and naive that I believed aga was the bees knees (even though this was post a1200/a4k/etc.).
-
Anyone else remember that aga was to be intergrated into the cpu die?
Somewhere in the back of my head I seem to remember that C= had actually got a licence for 680x0. I remember some passed-on posting from CIX about a new cost-reduced model that never saw the day of light.
-
If it helps, the Atari ST chips are 3um. I guess the Amiga ones are about the same.
/MikeJ
-
Anyone decapped an OCS chipset to see the die sizes and get some sexy images online?
Wikipedia says that an 8500 transistor chip (Z80) on 4um was 18mm^2 (just over 4mm by 4mm. The 8086 was a 3um chip and had 29,000 transistors on a slightly larger die. The 68000 was a 4um chip, but it looks like it had denser logic than the Z80, partially offsetting the extra transistor count (or it doesn't contain 68000 transistors).
-
Only the photographic equipment to get the image remains then .. ;)
> 3 µm should be doable.
-
Only the photographic equipment to get the image remains then .. ;)
> 3 µm should be doable.
Just ask the guys from http://www.visual6502.org.
They are done with the 6502 and 6800 chips. They have the 68000, the VIC-II, the SID, the TIA and others chips as WIP.
Regards,
Frederic
-
Have stuff, won't do. Is probably the situation ;)
-
We are making progress, but slowly. We've got many chips (http://visual6502.org/wiki/index.php?title=Chips_in_our_collection), but there are bottlenecks at deprocessing, at photography and at layout capture. And sometimes real life intrudes.
I stumbled across this thread (http://www.cpu-world.com/forum/viewtopic.php?p=112652#112652) some weeks ago which talks about photographic techniques, and links to some good z80 images too.
Cheers
Ed
-
This thread has made me thrawl the net for hours now. Even ordered a book too!
Great stuff. Sometimes I get to think that this internet lark was a good thing.
(Did those A8 chip schematics get posted though?)
-
We are making progress, but slowly. We've got many chips (http://visual6502.org/wiki/index.php?title=Chips_in_our_collection), but there are bottlenecks at deprocessing, at photography and at layout capture. And sometimes real life intrudes.
I stumbled across this thread (http://www.cpu-world.com/forum/viewtopic.php?p=112652#112652) some weeks ago which talks about photographic techniques, and links to some good z80 images too.
Cheers
Ed
Hi Ed, I assume you are part of the Visual6502 project? Wonderful idea! I hope you can hurry along the decapping guys as we all want to see the Amiga chips in their full glory... actually imagine a visual sim of the entire Amiga!!! ;)
-
Hi Ed, I assume you are part of the Visual6502 project? Wonderful idea! I hope you can hurry along the decapping guys as we all want to see the Amiga chips in their full glory... actually imagine a visual sim of the entire Amiga!!! ;)
Yes, I'm part of the team. It's become a slightly distributed effort for the polygon capture, but deprocessing and photography is especially skilled work and is still one guy.
Putting together a system simulation is an attractive idea - noting that the clock speed may be a million times slower than reality, so video frames could be an overnight wait - but it's most likely that some eight-bit system or other will appear before the Amiga, and neither will be soon, because of the bottlenecks. (Quietust is doing well with the NES (http://http://www.qmtpro.com/~nes/chipimages/#rp2c02), and Greg had an early demo of a partial 2600 - see page 44 of the pdf (http://visual6502.org/docs/6502_in_action_14_web.pdf))
Die photos of the decapped chip is always the first thing to appear though, and you'll see it on the blog (http://blog.visual6502.org/) as soon as it happens! We do have a chipset for an Amiga 500, and will get to it, but we've other promises to keep up first.
(Thanks to everyone who has donated chips!)
Cheers
Ed
-
Yes, I'm part of the team. It's become a slightly distributed effort for the polygon capture, but deprocessing and photography is especially skilled work and is still one guy.
Putting together a system simulation is an attractive idea - noting that the clock speed may be a million times slower than reality, so video frames could be an overnight wait - but it's most likely that some eight-bit system or other will appear before the Amiga, and neither will be soon, because of the bottlenecks. (Quietust is doing well with the NES (http://http://www.qmtpro.com/~nes/chipimages/#rp2c02), and Greg had an early demo of a partial 2600 - see page 44 of the pdf (http://visual6502.org/docs/6502_in_action_14_web.pdf))
Die photos of the decapped chip is always the first thing to appear though, and you'll see it on the blog (http://blog.visual6502.org/) as soon as it happens! We do have a chipset for an Amiga 500, and will get to it, but we've other promises to keep up first.
(Thanks to everyone who has donated chips!)
Cheers
Ed
I, for one, really can't wait to see those Amiga Chips stripped bare!
I think your work is the best way to preserve the effort that went into designing these computing devices. :)
-
I hope others try the decapping and photographing. Post-processing can be distributed among amiga fans.. ;)
-
Any decap progress? ;)
How does one interpretate an image like this (http://uxul.org/~noname/visual6502/8580/t/bf/20x/data/000031.png) anyway?
-
Any decap progress? ;)
How does one interpretate an image like this (http://uxul.org/~noname/visual6502/8580/t/bf/20x/data/000031.png) anyway?
Most of what you can see is wires. The little circles or dots on the wires are vias, which connect up a layer or down a layer to continue the wire on a different metal layer. Some of the large blocks may just be large chunks of metal wire. I don't see anything that strikes me as a transistor, but that may be lower stuff covered up by what we can see in the picture. It also helps to be able to turn on/off various layers to focus on different things. A static picture only lets you really see certain things. It helps to have done chip design to read this stuff. If you've done PCB design before you'll understand a lot of it as well.
-
Very cool stuff, this :) Wonder what it would cost to make new Amiga custom chips based on those images once completely reverse engineered.
-
Somewhere in the back of my head I seem to remember that C= had actually got a licence for 680x0. I remember some passed-on posting from CIX about a new cost-reduced model that never saw the day of light.
You want to read this: http://obligement.free.fr/articles_traduction/itwludwig_en.php
-
You want to read this: http://obligement.free.fr/articles_traduction/itwludwig_en.php
Good read, but this was when C= was still alive and was deffo a 68K chip in a cost reduced machine.
-
"so there's a 68000 compatible emulation mode for the PA-RISC." - cool!
Development path:
"We've had a lot of work spent on designing new instructions which allow us to have really fantastic 3D performance,"
(AAA + HP-PA-150 in a single chip)
Guess this why there's so little.. OMG-machines out there:
It's a matter of ensuring that there's some vision at the top of the company in order to let something like this go through. Commodore has too often in the past been in the situation of having designed very cool products which just at the very end of the life cycle after we've spent all the money on them don't actually make it to market because the people at the top have a lack of vision and when they see the finished product say "Oh no, they wouldn't wanna buy that." But I don't think we'll see that with the potential buyers.
The core business?
he best of the resources that Commodore's always had, basically designing our own chips rather than going out to someone else and just using an off-the-shelf chip.