@matthey: let's put it simple and completely ignore other factors than the lack of packed/chunky.

So, we can lift a bit the configurations and assume to have 6MB of Fast Mem added to the basic A1200, and a total of 8MB of Chip RAM to the "virtual" A1200.

The goal, for me, remains the same: check how competitive could have been the Amiga platform for 3D, when excluding the lack of packed/chunky graphics.

BTW and whilst 16-bit packed/chunky is visually very compelling, it would have doubled the assets space (which is mostly dominated by the textures), requiring much more memory. So, porting Doom to AA+ systems would have been possible, but required much more RAM.

@cdimauro

Henri Rubin has "little computer experience" and has a tech leadership position in a computer tech company.

Quote:

@cdimauro OK, so those are the numbers for first and last production period. Do you know how much costed the three chips for the A1200?

Quote:

From Commodore - The FInal Years, The current plans for the high-end computers rested on the AAA chipset. Unfortunately, Ted Lenthe predicted the chipset would cost $120 per system in low volumes or $60 in high volumes, a far cry from the $15 cost of the OCS developed by Jay Miner.

OCS for ~$15 vs AAA for $60.

_________________
Amiga 1200 (rev 1D1, KS 3.2, PiStorm32/RPi CM4/Emu68)
Amiga 500 (rev 6A, ECS, KS 3.2, PiStorm/RPi 4B/Emu68)
Ryzen 9 7900X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB

@cdimauro

Quote:

cdimauro wrote: @matthey: let's put it simple and completely ignore other factors than the lack of packed/chunky. So, we can lift a bit the configurations and assume to have 6MB of Fast Mem added to the basic A1200, and a total of 8MB of Chip RAM to the "virtual" A1200. The goal, for me, remains the same: check how competitive could have been the Amiga platform for 3D, when excluding the lack of packed/chunky graphics. BTW and whilst 16-bit packed/chunky is visually very compelling, it would have doubled the assets space (which is mostly dominated by the textures), requiring much more memory. So, porting Doom to AA+ systems would have been possible, but required much more RAM.

The problem with optimized Bittler C2P is the extra R&D time is figure out the software workaround while PC's VGA clones have "just works" chunky pixel hardware.

In Commodore's The FInal Years, Dave Haynie's Acutiator has a workaround with 68EC040.

Quote:

The engineers wanted to allow the then-new Motorola 68040 processor to work with the next generation of Amigas. And of course, the architecture would work with the upcoming AAA chipset, as well as the more imminent AA. And because AAA was designed to work with different processor families, Haynie wanted his Acutiator motherboard to also handle different processors. Specifically, there were at least three major RISC processor families at that time and he wanted Acutiator to accept these RISC chips. Their architecture required three custom chips: EPIC, AMOS, and SAIL. In cost comparisons, Haynie calculated that the Acutiator architecture would add approximately $125 to a system (including the cost of a 68EC040 chip), resulting in a $300 retail price increase. This was a bargain, considering the user received a significant processor upgrade. Haynie proposed that Commodore should assign Scott Shaeffer, Paul Lassa, and himself to each create the three required gate array chips. He expected prototypes in 7 to 9 months, with the first systems shipping in 1992.

68EC040-25 is about $100.

Without the new custom chips, a lower cost of 68EC040 would be out of reach for AA Amigas.

Last edited by Hammer on 15-Aug-2024 at 08:56 AM.

_________________
Amiga 1200 (rev 1D1, KS 3.2, PiStorm32/RPi CM4/Emu68)
Amiga 500 (rev 6A, ECS, KS 3.2, PiStorm/RPi 4B/Emu68)
Ryzen 9 7900X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB

@cdimauro

Quote:

cdimauro wrote: So, we can lift a bit the configurations and assume to have 6MB of Fast Mem added to the basic A1200, and a total of 8MB of Chip RAM to the "virtual" A1200.

Smart-ass question: what about the 4MB of memory space that is reserved for PCMCIA? Let's just say you can use the "ranger ram" address space, so 4 MB + 1.8 MB? How do you set this up physically? 4 MB of RAM on motherboard, 2 MB for chipram and 2 MB of which 1.8MB is used as "ranger ram", and then a SIMM socket for an optional 4MB chip? Gehh.... isn't it easier to user the trapdoor and have a card with full 020/030, simm socket and perhaps RTC? :p

Funny thing - CBM never made any expansion card for A1200, did they....

Last edited by kolla on 15-Aug-2024 at 09:40 AM.

_________________
B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC

@Hammer

Quote:

Hammer wrote: The problem with optimized Bittler C2P is the extra R&D time is figure out the software workaround while PC's VGA clones have "just works" chunky pixel hardware.

It's not the big deal you make it out to be. A 'trivial' c2p routine only takes minutes to create. Fully optimizing it might take a day or two - a minuscule effort compared to all the other stuff needed in a typical game. Doom took over a year to develop.

Quote:

Their architecture required three custom chips: EPIC, AMOS, and SAIL. In cost comparisons, Haynie calculated that the Acutiator architecture would add approximately $125 to a system (including the cost of a 68EC040 chip), resulting in a $300 retail price increase. This was a bargain, considering the user received a significant processor upgrade... He expected prototypes in 7 to 9 months, with the first systems shipping in 1992. 68EC040-25 is about $100.

I'm skeptical. Who was going to make these 3 chips for $25?

And EC040 sucked. Power hungry but no FPU? No thanks.

@Hammer

Quote:

Hammer wrote: @cdimauro Henri Rubin has "little computer experience" and has a tech leadership position in a computer tech company.

Jack Tramiel as well, but he was the one to make Commodore great...
Quote:

Quote: @cdimauro OK, so those are the numbers for first and last production period. Do you know how much costed the three chips for the A1200? Quote: From Commodore - The FInal Years, The current plans for the high-end computers rested on the AAA chipset. Unfortunately, Ted Lenthe predicted the chipset would cost $120 per system in low volumes or $60 in high volumes, a far cry from the $15 cost of the OCS developed by Jay Miner. OCS for ~$15 vs AAA for $60.

OK, but that's only AAA. Not AA.

AA used 1.5u process for Alice and Lisa, and used not so many more transistors compared to OCS. So, probably it costed less than OCS.

@Hammer

Quote:

Hammer wrote: @cdimauro Quote: cdimauro wrote: @matthey: let's put it simple and completely ignore other factors than the lack of packed/chunky. So, we can lift a bit the configurations and assume to have 6MB of Fast Mem added to the basic A1200, and a total of 8MB of Chip RAM to the "virtual" A1200. The goal, for me, remains the same: check how competitive could have been the Amiga platform for 3D, when excluding the lack of packed/chunky graphics. BTW and whilst 16-bit packed/chunky is visually very compelling, it would have doubled the assets space (which is mostly dominated by the textures), requiring much more memory. So, porting Doom to AA+ systems would have been possible, but required much more RAM. The problem with optimized Bittler C2P is the extra R&D time is figure out the software workaround while PC's VGA clones have "just works" chunky pixel hardware.

In fact, I want to put the C2P out of my analysis.
Quote:

In Commodore's The FInal Years, Dave Haynie's Acutiator has a workaround with 68EC040. Quote: The engineers wanted to allow the then-new Motorola 68040 processor to work with the next generation of Amigas. And of course, the architecture would work with the upcoming AAA chipset, as well as the more imminent AA. And because AAA was designed to work with different processor families, Haynie wanted his Acutiator motherboard to also handle different processors. Specifically, there were at least three major RISC processor families at that time and he wanted Acutiator to accept these RISC chips. Their architecture required three custom chips: EPIC, AMOS, and SAIL. In cost comparisons, Haynie calculated that the Acutiator architecture would add approximately $125 to a system (including the cost of a 68EC040 chip), resulting in a $300 retail price increase. This was a bargain, considering the user received a significant processor upgrade. Haynie proposed that Commodore should assign Scott Shaeffer, Paul Lassa, and himself to each create the three required gate array chips. He expected prototypes in 7 to 9 months, with the first systems shipping in 1992. 68EC040-25 is about $100. Without the new custom chips, a lower cost of 68EC040 would be out of reach for AA Amigas.

It was too expensive. Also, why waste time to support also RISCs? Bah...

@kolla

Quote:

kolla wrote: @cdimauro Quote: cdimauro wrote: So, we can lift a bit the configurations and assume to have 6MB of Fast Mem added to the basic A1200, and a total of 8MB of Chip RAM to the "virtual" A1200. Smart-ass question: what about the 4MB of memory space that is reserved for PCMCIA?

PCMCIA = 16-bit interface = crappy slow -> not an option.
Quote:

Let's just say you can use the "ranger ram" address space, so 4 MB + 1.8 MB? How do you set this up physically? 4 MB of RAM on motherboard, 2 MB for chipram and 2 MB of which 1.8MB is used as "ranger ram", and then a SIMM socket for an optional 4MB chip? Gehh.... isn't it easier to user the trapdoor and have a card with full 020/030, simm socket and perhaps RTC? :p

In fact, I was talking about the regular A1200 and nothing else. Memory expansion -> trapdoor.

But if you want to talk about a change on the A1200, then I would have removed the socket for the FPU and used the space & cost for a SIMM socket to expand memory: cheap, easy, and effective (and the trapdoor which could still be used for accelerators or other things).
Quote:

Funny thing - CBM never made any expansion card for A1200, did they....

Others did it, which was fine.

@cdimauro

Quote:

cdimauro wrote: PCMCIA = 16-bit interface = crappy slow -> not an option. ... In fact, I was talking about the regular A1200 and nothing else. Memory expansion -> trapdoor.

Well, the "regular A1200" _does_ have PCMCIA, and _does_ have memory address range reserved for it, so in principle you cannot add more than 4MB of fast RAM on a stock A1200 (or A600, the 68000 and the 68EC020 on Amiga sharing the "23 bit" address space). If you do, you must be clever and either ensure PCMCIA is disabled, or do some more trickery (patch drivers etc) to ensure used pcmcia address space doesn't collide with fast RAM addresses.

Last edited by kolla on 22-Aug-2024 at 06:29 PM.

_________________
B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC

@kolla

For me, the PCMCIA slots are used for cnet.device based NIC. They might not have been the fastest solutions but I can't overstate how useful they have been.

_________________
Doing stupid things for fun...

@Karlos

Same here, and yes.
Just making the point that simply adding 6 MB of fast ram on card may not be as trivial as it sounds.

_________________
B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC

@kolla

Quote:

kolla wrote: @cdimauro Quote: cdimauro wrote: PCMCIA = 16-bit interface = crappy slow -> not an option. ... In fact, I was talking about the regular A1200 and nothing else. Memory expansion -> trapdoor. Well, the "regular A1200" _does_ have PCMCIA, and _does_ have memory address range reserved for it, so in principle you cannot add more than 4MB of fast RAM on a stock A1200 (or A600, the 68000 and the 68EC020 on Amiga sharing the "23 bit" address space). If you do, you must be clever and either ensure PCMCIA is disabled, or do some more trickery (patch drivers etc) to ensure used pcmcia address space doesn't collide with fast RAM addresses.

OK, now I got what you were talking about. Right, you've to disable the PCMCIA if you want to have more than 6MB of memory.

Or, you have to resort to something like 4MB of Fast mem + 1,75MB of "Slow" (as address range) mem, which is tricky.

I don't know if the PCMCIA address space could be reduced to 2MB, for example.

@bhabbott

Quote:

@bhabbott It's not the big deal you make it out to be. A 'trivial' c2p routine only takes minutes to create. Fully optimizing it might take a day or two - a minuscule effort compared to all the other stuff needed in a typical game. Doom took over a year to develop.

1. Doom started its development after Wolf3D.

John Carmark's ShadowCaster 3D engine is between Wolfenstein 3D and Doom 1. https://en.wikipedia.org/wiki/ShadowCaster


2. https://bigbookofamigahardware.com/bboah/product.aspx?id=1604
Quote:

The “chunky to planar� logic was thought out in a lunchtime conversation between Beth Richard (system chip design), Chris Coley (board design), and Ken Dyke (software) over Subway sandwiches on a picnic table in a nearby park one day, because Ken was telling us how much of a pain it was to shuffle bits in software to port games from other platforms to the Amiga planar system. We took the idea to Hedley Davis, who was the system chip team manager and lead engineer on Akiko and he said we could go ahead with it. I showed him the “napkin sketch� of how I thought the logic would work and was planning on getting to it the next day as it was already late afternoon by that point. I came in the next morning and Hedley had completed it already, just from the sketch!

Commodore software engineer, Ken Dyke's arguments were repeated by John Carmack in 1994.

EA's Bull Frog didn't complete A1200's Magic Carpet R&D.

Your assumptions are flawed.

Quote:

@bhabbott I'm skeptical. Who was going to make these 3 chips for $25?

The cited CSG chip prices are typical. You don't properly read "Commodore - The Final Year" book.

These are glue support chips for 68EC040 and Amibus (Amiga chipset side).

Quote:

@bhabbott And EC040 sucked. Power hungry but no FPU? No thanks.

Sony's PSX deserves to win with the likes of you.

Last edited by Hammer on 24-Aug-2024 at 08:11 AM.

_________________
Amiga 1200 (rev 1D1, KS 3.2, PiStorm32/RPi CM4/Emu68)
Amiga 500 (rev 6A, ECS, KS 3.2, PiStorm/RPi 4B/Emu68)
Ryzen 9 7900X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB

@cdimauro

Quote:

In fact, I want to put the C2P out of my analysis.

That's your choice.

Quote:

It was too expensive. Also, why waste time to support also RISCs? Bah..

"Cheapo RISC".

For 1992, the three big-endian capable RISC chip examples are MIPS, PA-RISC, and PowerPC.

For $100 in 1992, it's 386DX-25 from Intel. LOL

It's good that the "human bus error" (Bill Sydnes) and pro-PC (Jeff Franks) in Commodore didn't allow 68EC040-25 powered Amiga during $100 386DX-25 to 386DX-33 era. LOL

It's good that pro-PC Bill Sydnes and Jeff Franks bankrupted Commodore into oblivion i.e. "Insidious Inside".

_________________
Amiga 1200 (rev 1D1, KS 3.2, PiStorm32/RPi CM4/Emu68)
Amiga 500 (rev 6A, ECS, KS 3.2, PiStorm/RPi 4B/Emu68)
Ryzen 9 7900X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB

@kolla

Quote:

kolla wrote: @cdimauro Quote: cdimauro wrote: PCMCIA = 16-bit interface = crappy slow -> not an option. ... In fact, I was talking about the regular A1200 and nothing else. Memory expansion -> trapdoor. Well, the "regular A1200" _does_ have PCMCIA, and _does_ have memory address range reserved for it, so in principle you cannot add more than 4MB of fast RAM on a stock A1200 (or A600, the 68000 and the 68EC020 on Amiga sharing the "23 bit" address space). If you do, you must be clever and either ensure PCMCIA is disabled, or do some more trickery (patch drivers etc) to ensure used pcmcia address space doesn't collide with fast RAM addresses.

From https://www.amigakit.com/a1200-fast-memory-expansion-40mhz-option-p-12690.html
For A1200, I have 8 MB Fast RAM expansion card from Amigakit which allows up to 5.5 MB Fast RAM with PCMCIA compatible. 5.5 MB mode requires a software patch i.e. 1.5MB of $C0 Ranger memory address range is reused for this expansion card.

I enabled $C0 Ranger memory on Emu68 to improve backward compatibility.

My other Wicher 508i accelerator for A500 with 1 MB Chip RAM can also reuse $C0 Ranger memory range. Wicher 508i's $C0 usage doesn't need a software patch in the start-up sequence.

Certain Amiga games are brain-dead looking for $C0 Ranger memory.

Last edited by Hammer on 24-Aug-2024 at 08:38 AM.

_________________
Amiga 1200 (rev 1D1, KS 3.2, PiStorm32/RPi CM4/Emu68)
Amiga 500 (rev 6A, ECS, KS 3.2, PiStorm/RPi 4B/Emu68)
Ryzen 9 7900X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB

@Hammer

Quote:

Hammer wrote: @cdimauro Quote: In fact, I want to put the C2P out of my analysis. That's your choice.

It's not my choice: you haven't understand the precise context (as usual).
Quote:

Quote: It was too expensive. Also, why waste time to support also RISCs? Bah.. "Cheapo RISC". For 1992, the three big-endian capable RISC chip examples are MIPS, PA-RISC, and PowerPC. For $100 in 1992, it's 386DX-25 from Intel. LOL It's good that the "human bus error" (Bill Sydnes) and pro-PC (Jeff Franks) in Commodore didn't allow 68EC040-25 powered Amiga during $100 386DX-25 to 386DX-33 era. LOL It's good that pro-PC Bill Sydnes and Jeff Franks bankrupted Commodore into oblivion i.e. "Insidious Inside".

RISCs, RISCs, always RISCs. If you like RISCs, go for them! Amiga was NOT a platform for RISCs.

And I reveal you a secret: the above PCs that you've mentioned... have NOT used RISCs.

So, why on Earth the Amiga should have used a RISC?!? There was no need for it: we had very good processors at the time. What was really missing is a DECENT evolution of the hardware, and this is your heroes' responsibility.