Unreleased Amiga chipset better than anything today?

[aeberbach]

Originally Posted by Bruce Abbott

Nope. You couldn't squeeze 2MB into the A1000 in 1985. It would need a much bigger case.

I had a Phoenix 2MB expansion for my A1000, connecting to the right side expansion. Of course it went outside the main case in a chunky steel enclosure but there was room inside the A1000 where that expansion's PCB could have gone. It was just a ridiculous amount of RAM at design time so nobody bothered to allow for it - nobody would ever need or be willing to pay for that much memory.

[Promilus]

Originally Posted by Bruce Abbott

Nonsense, it had everything to do with it. Cost was a primary factor in the design of the Amiga. Everything was built around a shared memory architecture that made the best possible use of limited RAM. It's the primary reason the Amiga was able to do so much at such a low price.

2MB chipram space was reserved pretty much from the start. They just made limited AGNUS versions and due to IC specifics didn't include option to go beyond 512K on A1000 and rev3, rev5 A500 and beyond 1MB on rev6 A500 (actually there's no option to get 1MB on rev6 A500 without adjustments to the board - there's room for whole 1MB on board which shows it was designed to hold that much there but it WAS NOT introduced until much later with A500+)

Nope. You couldn't squeeze 2MB into the A1000 in 1985. It would need a much bigger case.

Not really. 30 pin SIMM was already invented... You could invent own slot and stack up RAM modules. But don't tell me it was impossible to get 2MB of chipram. Sure it was. Impossible is to get 256GB DDR4 RAM for desktop PC nowadays despite the fact CPU can address that much. Why? Because that would require 4x64GB modules and the max is 16GB iirc (32 for DDR5). So given that - A1000 has artificial limitations which were dropped down in ECS (and yes, A1200 was supposed to have only 1MB of chipram on stock machine as well!). Also - 2 parts didn't ever get any chance to upgrade during whole amiga lifetime - Paula and CIA.

It was not beside the point. The A1000 was cheaper than the IBM PC jr (their 'low cost' home computer for gaming etc.) released a year earlier. The A1000's price was the upper end of what home computer users were willing to spend. Another $500 would have made it unsaleable.

It is because A1000 while cheaper did absolutely destroy much more expensive PCs around the same age. A1200 did not.

How much did most other home computer platforms advance?

C64: no improvement in 12 years.

Except C128 model which you can hardly call an improvement (ok, floppy is much better and basic version is also good) and C65 which was not only in development but rather nice number of prototypes were already built and were fairly working, unlike AAA...

ZX Spectrum: No graphics improvement in 9 years

it was always cheap ass solution.

Amstrad CPC: no improvement in 6 years.

I did not follow Amstrad products anyway so I won't be commenting that.

Acorn Archimedes: no improvement in 9 years.

There were next models in the lineup up to RiscPCs so get your damn facts straight.

The PC started with mono text and the Mac mono 2 color graphics, so both had a long way to go to catch up with the Amiga. The ST never really made use of the few improvements it got.

From 85 where Amiga came out I think it's pretty obvious to any non colorblind weirdo PC did outtake Amiga in development speed of new standards. EGA 84, VGA 87 and enhanced VGA afterwards. So it did suffer from poor isa throughput. VLB - 1992, it destroys everything AGA - released the same year - can throw at it. I rest my case.

But why did most home computer platforms take so long to be improved if at all?
Had Commodore produced a radically new chipset every few years it would have split the user base and made software development and distribution more complicated.

They already did that and in not particularly good moment. And they were planning to do it even further but died before it came to reality. So - again - your argument makes no sense when you take into account what Commodore did and was planning to do.

What you call "did not improve" I call providing stability.

What you call stability I call stagnation. That's something which was tolerable in consoles. It was not in computers.

No, it wasn't. A typical 386SX with ISA VGA card in 1992 did not beat up the Amiga 1200.

that's fvking old tech you biased idiot! You are always doing that, comparing latest released Amiga with some dusty cheap PC most ppl did buy exactly because it was cheap outdated shit and not latest tech!

Even in 1992 you couldn't buy any home computer that came close to what the Amiga could do for the price. Which was the whole point.

It makes absolutely no sense IF YOU CAN AFFORD a computer which does MUCH MORE at the same time. It was IMPOSSIBLE in 85. It surely was possible in 92. So again - get your FACTS straight.

You are trying to say that the Amiga was light-years ahead of anything else in 1985, so you think it should still have been light-years ahead of everything else in 1992.

Well that's the actual point of whole Moore's law debate, did you guessed that just now?

But that's not how technology works.

Sure it does if one wants to keep dominance in some area. If development becomes stagnant you lose the lead and the market. It was absolutely shitty idea to buy Amiga chipset, hook it up as computer, not update design regularly with new models and then just throw it out of the window with new design when current one after minor alterations can't be even sold as cheap ass home computer.

Why do Amiga fans get so upset about this? My guess is that their egos were built on the Amiga being superior to every other platform, and when it eventually didn't they felt inadequate - and needed someone to blame.

Well if their egos was built that way by Commodore itself. So pin that blame on them...

[hammer]

Originally Posted by Promilus

2MB chipram space was reserved pretty much from the start. They just made limited AGNUS versions and due to IC specifics didn't include option to go beyond 512K on A1000 and rev3, rev5 A500 and beyond 1MB on rev6 A500 (actually there's no option to get 1MB on rev6 A500 without adjustments to the board - there's room for whole 1MB on board which shows it was designed to hold that much there but it WAS NOT introduced until much later with A500+)

A500 Rev6A has 2MB Chip RAM jumpers.

For A500 Rev6A's 2MB Chip RAM config, the four RAM modules need to be changed, 2MB Agnus ECS (e.g. 8372B), and jumpers (JP2, JP3, JP5, and JP7A).

Read https://www.amibay.com/threads/2megs...erboard.89654/
for instructions for A500 Rev 6A's 2 MB Chip RAM configuration.

A500 Rev6 PCB has a reserved 2MB Chip RAM design.

[hammer]

Originally Posted by Promilus

Stock A1200 doesn't have fast ram. Actually no stock "desktop" Amiga has.
.

FALSE. My "desktop" Amiga 3000 with Kickstart 2.04 ROM has 1 MB Chip RAM and 1 MB Fast RAM as stock configuration.

Commodore expected the end user to move 1 MB Fast RAM chips into the Chip RAM location when expanding Fast RAM.

From Commodore Australia's Amiga 3000 configuration from July 1990.

https://archive.org/details/Australi...p?view=theater


From The Australian Commodore and Amiga Review, July 1990.

This 1990 baseline Amiga 3000 configuration has 1 MB Chip RAM and 1 MB Fast RAM as stock configuration.

A3000's Flicker Ficker made 640x400i/512i 16 colors flicker-free and faster 32-bit Chip RAM to be roughly on par with 1987 era VGA's 640x480 16 colors. A3000's ECS is faster than A500's ECS due to less memory bottleneck.

My A3000 didn't have the early A3000 SuperKickstart 1.4 config since it has Kickstart 2.04 ROM. I later updated Kickstart ROM into release 3.1 and 4MB ZIP Fast RAM. I obtained my ex-corporate A3000 in early 1992.

[hammer]

Originally Posted by Bruce Abbott

This would be for lack of will, not ability.

The right thing to do was develop a new API that provided fully virtualized access to the graphics hardware. That's what EGS did. You want your application to use the new RTG graphics? recompile your code! That's what PC developers were doing, and Amiga developers would have too. It would create more sales!

But Commodore was stuck on maintaining backwards compatibility.

For 1992, GVP's EGS-28/24 Spectrum wasn't cost-effective compared to PC's $169 USD ET4000AX running with Windows 3.1.

GVP Spectrum in 1994 Canada has a $625.00 asking price.

In 1992, GVP was one of the strong 3rd party Amiga add-on vendors that couldn't match Diamond Multimedia Inc.'s economies of scale.

https://archive.org/details/bub_gb_h.../n603/mode/2up
PC Mag 1992-08, page 604 of 664,
Diamond Speedstar 24 (ET4000AX ISA) has $169 USD.

Trident 8900CL's performance is close to Diamond Speedstar 24 (ET4000AX ISA).

The major problem is the "Amiga Tax".

For 1992, the Amiga has lost its "power without the price" advantages.


[ Show youtube player ]

Doom (low details) on 386DX-40 with 128K cache
Tseng ET4000 ISA = 26.751 fps (1989 release graphics card)
Trident 8900CL ISA = 23.0088 fps
WD90C32 = 26.838 fps (Diamond Speedstar 24X)

They perform like A1200 AGA with 68030 @ 40 Mhz to 50 Mhz! A1200 arrived in Q4 1992 with limited numbers and a large price hole between A1200 and A4000/040.

Originally Posted by Bruce Abbott

The main advantage of VGA was that all screen modes used 31kHz or above, including emulation of CGA and EGA 15kHz modes. This created a unified standard and opened the door for future enhancements like SVGA (800x600 in 16 colors).

The VGA standard has a 31 kHz refresh rate promotion for legacy EGA and CGA modes.

A3000 has an expensive Amber flicker fixer with a 1 MB frame buffer.

"Read My Lips, No New Chips" during A3000's development.

Originally Posted by Bruce Abbott

The A3000 shouldn't even have existed. But somebody thought Commodore should get into workstations, where the emphasis was on high resolution for CAD etc., not gaming. AGA was slow coming because they tried to make AA too high-end.

A3000's CPU has a 32-bit memory controller chip on 32-bit Chip RAM that co-existed with 16-bit Agnus ECS.

C65's 256-color (8-bit planes) display-capable chipset was completed in December 1990 and it was canceled. This is R&D resource wastage.


Commodore management memo "Read My Lips, No New Chips" during A3000's development.



AAA's AGA fork was completed in Feb 1991 and is operational for A3000+ AGA.

https://bigbookofamigahardware.com/b...uct.aspx?id=23
A3000+ According to Dave Haynie

"The Amiga 3000+ was the first computer based on the Pandora chipset (which was later dubbed AA, then AGA). Revision 0 of this system first booted successfully in February of 1991, thanks due to a chip revision that got the display logic actually working. This is revision 1, which completed the audio subsystem, and moved to surface-mount components.

As the name suggests, this system was being designed as a drop-in replacement for the existing Amiga 3000 motherboard. In addition to the features everyone knows from the Amiga 3000, and from later computers such as the Amiga 4000, the Amiga 3000+ sported the AT&T DSP3210 digital signal processor as a coprocessor.

You can see quite a bit of support circuitry for the DSP in the upper lefthand corner of this board. There was an audio CODEC here, designed to allow 16-bit, 2-channel recording and playback. This was very cutting edge at the time, such chips, common today, where just becoming available. In addition, there was a separate mono CODEC with hardware phase correction, which supported modem protocols up to V32. The actual DSP was located above and to the right of the CPU.

Note: this is the real Amiga 3000+, very rare. Most of the Amiga 3000+ type systems out, whether boards or whole computers, are actually the scaled-down "AA3000", which was after the A3000+ had been cancelled, by the Bill Sydnes administration, as a product." - Dave Haynie


A3000+ AGA's cancellation delayed the AGA A1200 project.

This "A3000+ AGA" turned into an A4000 after project reactivation and its cost-reduced version is the A1200.

A1200 includes a cut-down Buster chip known as Budgie which includes a 32bit Fast RAM memory controller.

Dave Haynie is aware math compute power problem for the AGA chipset, hence AT&T DSP3210 (Dave Haynie's RISC SuperFX tactic).

The problem is Commodore's upper management.

Lost time for AAA.
Lost time for AA/AGA.


Commodore's leadership list and their exits
Jack Tramiel, exited in 1984.

Marshall Smith, exited in 1986.

Thomas Rattigan, exited in 1987(?).

Max Toy, exited in late 1988-1989(?). This person is a problem.

Mehdi Ali became Commodore International's president early in 1989. This person is a problem.

[AestheticDebris]

Originally Posted by Bruce Abbott

Amstrad CPC: no improvement in 6 years.

I'd give you the others, but the CPC gained a considerable graphics upgrade within 5 years. Upping the palette from 27 to 4096 colours, adding hardware sprites and pixel scrolling. And it is still universally decried as "too little, too late"

The mindset that manufacturers "didn't need" to upgrade the specs of machines was half the reason the PC succeeded. If anyone else had their eye on the ball, they'd have seen how quickly third party expansions were improving and found a way to take advantage of them. Focusing too much on making their own chipset (and taking far too long over it) hampered Commodore, at least partly because they were stuck in the mindset that the Amiga could go on for as long as the C64.

[hammer]

Originally Posted by pandy71

Don't get me wrong but everything listed by you is not Amiga - emulators based on UAE and WinUAE are just Amiga emulators. AROS is not Amiga is a Amiga OS recreation independently from HW.
Hombre was not planned with x86 as CPU.

Commodore's Dr. Ed Helper rejected PowerPC. Look in the mirror.

AROS is not limited by the CPU instruction set family.

Native MoprhOS 3.12 beta runs on AMD64 [ Show youtube player ]

Originally Posted by pandy71

And you can achieve this in many ways - but this is not CF issue.

The issue is the approach with hosting the legacy CPU layer i.e. Apple's userland's 68K/X86 emulator to RISC vs. X86 PC's transparent "hypervisor" X86 translator to RISC.

FireTOS implemented a 68K emulator via Apple's userland method.

Any good RISC CPU can host the "Emu68" method.

Originally Posted by pandy71

Don't forget about FPGA Amiga installed base...

There are more than 10,000 AC68080s and an unknown number of MiniMig, MiSTer/MIST (Amiga 68K configured).

[pandy71]

Originally Posted by hammer

A3000 has an expensive Amber flicker fixer with a 1 MB frame buffer.

A3000 is equipped with 1MB frame buffer? Where? Source please!

Originally Posted by hammer

Commodore's Dr. Ed Helper rejected PowerPC. Look in the mirror.

AROS is not limited by the CPU instruction set family.

Native MoprhOS 3.12 beta runs on AMD64 [ Show youtube player ]

AROS or MorphOS are not Amiga - it's exactly the same as with Amix or custom OS that never happened to compete with ST on MIDI area where multitasking hampered MIDI and introduced latency.

Originally Posted by hammer

The issue is the approach with hosting the legacy CPU layer i.e. Apple's userland's 68K/X86 emulator to RISC vs. X86 PC's transparent "hypervisor" X86 translator to RISC.

FireTOS implemented a 68K emulator via Apple's userland method.

Any good RISC CPU can host the "Emu68" method.

Of course and CF is not bad either but simply not good as expected to be.

Originally Posted by hammer

There are more than 10,000 AC68080s and an unknown number of MiniMig, MiSTer/MIST (Amiga 68K configured).

And? Please return to me when AC68080 will be available as ASIC ready to be bought same as RP2040 or some RISC V.
Price for AC68080 is too high when compared to performance.

[Bruce Abbott]

Originally Posted by aeberbach

I had a Phoenix 2MB expansion for my A1000, connecting to the right side expansion. Of course it went outside the main case in a chunky steel enclosure but there was room inside the A1000 where that expansion's PCB could have gone. It was just a ridiculous amount of RAM at design time so nobody bothered to allow for it - nobody would ever need or be willing to pay for that much memory.

I had a Microbotics Starboard 2 on my A1000. I bought it second-hand in 1988. it had:-

• 32 256k×1, 150 ns DIPs for 1 MB RAM (on "Main Deck")
• optional "Upper Deck" module with 32 DIPs for the other 1 MB

A total of 64 256kx1 DRAM chips in 16 pin DIP packages. That certainly wouldn't fit on the A1000 motherboard. It could have gone on a daughterboard, except that the 256k WCS board was taking up half the available space.

Of course as you say, 2.5MB was a huge amount of RAM for a home computer back in 1984 when the A1000 was designed. The IBM PC-AT, introduced in 1984, only had a maximum of 512kB on the motherboard.

[Bruce Abbott]

Originally Posted by Promilus

2MB chipram space was reserved pretty much from the start. They just made limited AGNUS versions...

Originally the base model was going to have 128k, but by the time it was finished they (sensibly) decided to go for 256k with another 256k in the expansion bay. So while a 2MB space was allocated in the memory map, nobody was thinking it would be filled up any time soon. The cost would be prohibitive.

...and due to IC specifics didn't include option to go beyond 512K on A1000

The A1000 used a 48 pin DIP Agnus chip. This was already larger than the maximum of 40 pins that most chips used. Going to PLCC in the A500 gave them more pins to play with.

and rev3, rev5 A500 and beyond 1MB on rev6 A500 (actually there's no option to get 1MB on rev6 A500 without adjustments to the board - there's room for whole 1MB on board which shows it was designed to hold that much there but it WAS NOT introduced until much later with A500+)

Yep. Each revision was looking towards further expansion. Commodore tried to keep their options open depending on where the DRAM market went. In 1988 there was a big shortage due to the US embargo on Japanese chips, which was triggered by a glut in 1985 when they made too many 256k chips and 'dumped' them on the market.

A1000 has artificial limitations which were dropped down in ECS (and yes, A1200 was supposed to have only 1MB of chipram on stock machine as well!).

All limitations are 'artificial'. I bought an Amstrad CPC664 with 64k in 1985. A mere 6 months later Amstrad released the CPC6128 which had 128k (using 16 64kx1 DRAMs). I responded by unsoldering the 64kx1 chips from my 664 motherboard and replacing them with 256kx1 chips, and made a compatible bank switching circuit to control them. 4 times the RAM in the same space!

Also - 2 parts didn't ever get any chance to upgrade during whole amiga lifetime - Paula and CIA.

They didn't need upgrading.

It is because A1000 while cheaper did absolutely destroy much more expensive PCs around the same age. A1200 did not.

As I said before, you can't expect the same from a maturing technology. In 1985 the Amiga managed to fill a niche others hadn't fully exploited with an innovative design. By 1992 several companies were doing nothing but designing graphics chips. They are still doing it today and no computer manufacturer would think of trying to beat them at that game.

The A1000 only 'destroyed' PCs in some areas. In others it was weak. No text mode, no hires non-interlace, poor performance in 16 color hires. Sure it had great hardware for games which PCs didn't, but that's because PCs weren't designed for games. Later, when they were, the Amiga would not be able to beat those much more expensive PCs.

Except C128 model which you can hardly call an improvement

That's why I didn't mention it. The C64 lasted longer than the C128, which was never designed to last very long. The C128 was a horrible kludge using the graphics chip from their failed C900 design, and a Z80 jammed into it to do CP/M.

The successor to the C64 would have been the C65. This was quashed for being too much like the Amiga but not as good - one example of Commodore of acting sensibly and us being the poorer for it!

There were next models in the lineup up to RiscPCs so get your damn facts straight.

Launched in 1994, 7 years after the Archimedes.

From 85 where Amiga came out I think it's pretty obvious to any non colorblind weirdo PC did outtake Amiga in development speed of new standards. EGA 84, VGA 87 and enhanced VGA afterwards. So it did suffer from poor isa throughput. VLB - 1992, it destroys everything AGA - released the same year - can throw at it. I rest my case.

In 1992 VLB was only available in high-end 486 machines. The Amiga had a 32 bit bus since 1990 with the A3000, and graphics cards that used it started appearing in 1992. By 1993 when VLB was becoming popular there were several cards for the A3000/4000 using the same chipsets as PC VL bus cards. They were expensive for sure, but if price wasn't a concern...

OTOH if price was a concern then the A1200 beat the pants off a typical 386SX that cost a lot more.

Let's get real here. The Amiga was originally designed to be a home computer like the C64 and Atari 800 etc., with best in class hardware and cheap enough that anyone could afford it. Eventually, when PCs got powerful enough to play games well too, people decided they could afford the extra money for a PC - and that was the end of the Amiga. The chipset was never designed for high-end stuff and trying to make it so was a waste of time, as Commodore's engineers eventually realized.

But AGA did give the A1200 better graphics than anything else in its class. Those of us who appreciated the Amiga didn't fret about what was coming out in the PC world, we simply enjoyed the improvements AGA provided. 30 years later we are still enjoying it, with developments like Vampire and PiStorm providing enough computing power to make AGA do stuff we couldn't even imagine in 1992.

What you call stability I call stagnation. That's something which was tolerable in consoles. It was not in computers.

I have to disagree. Other home computers 'stagnated' just as much, and fans didn't complain because they liked that stability. When the A1200 came out many A500 users argued that their existing machines were good enough.

that's fvking old tech you biased idiot! You are always doing that, comparing latest released Amiga with some dusty cheap PC most ppl did buy exactly because it was cheap outdated shit and not latest tech!

Funny. I have a New Zealand PC magazine from 1994 with an article arguing that buying the latest tech was not a good idea. They suggested buying a 386SX in 1992 and waiting until 486's got cheaper. The example showed significant savings to a business that did that.

It makes absolutely no sense IF YOU CAN AFFORD a computer which does MUCH MORE at the same time. It was IMPOSSIBLE in 85. It surely was possible in 92. So again - get your FACTS straight.

In 1985 I wanted an Amiga. But they weren't available in New Zealand so I would have to import a 120V NTSC model from the US at great expense. I bought an Amstrad CPC664 instead. I had lots of fun with that for a couple of years, then I got an A1000 in 1987. I still used the CPC for while after that too! Then I got an IBM Jx (Japanese version of the PC jr) when it was being dumped for $1000, and had lots of fun with that too - even though it wasn't nearly as powerful as the Amiga. My only 'mistake' was buying an A3000 for NZ$7200 in 1991. I could afford it, but it wasn't worth the money.

Sure it does if one wants to keep dominance in some area. If development becomes stagnant you lose the lead and the market. It was absolutely shitty idea to buy Amiga chipset, hook it up as computer, not update design regularly with new models and then just throw it out of the window with new design when current one after minor alterations can't be even sold as cheap ass home computer.

You know what would have been shitty? Telling Amiga corp you weren't interested in their computer but might use the chipset in something else. That's what Atari offered them.

Commodore took Jay Miner's design and turned it into a real product that became famous. He said they did a wonderful job of that. Then they applied the same strategy that they did with their very successful C64 - refine the design and then keep selling it for as long as possible, just like other home computer makers did. Maybe you can fault them for that, but I won't judge. I'm sure everyone at Commodore was doing what they thought was best.

If it wasn't for Commodore's efforts there would be no Amiga. They were committed to it until the end. Your lack of appreciation for all the hard work the people at Commodore did in order to deliver wonderful Amiga models unto our hands shows who is being shitty.

[Bruce Abbott]

Originally Posted by pandy71

And? Please return to me when AC68080 will be available as ASIC ready to be bought same as RP2040 or some RISC V.
Price for AC68080 is too high when compared to performance.

I have a Vampire V2 in my A600. It's twice as fast as the A3000 I used to have with 50MHz 060, has 4 times the RAM, faster RTG, and is more compatible. And way way cheaper. I tried other accelerator cards in the A600 and they were underwhelming - not the Vampire. It's great for browsing the web and playing those ported games Duke Nuken 3D and Quake. Good for emulators too. I even managed to play Tomb Raider (slowly) under PC Task!

The latest V4 cards are a bit more expensive, but even more powerful and have more features. If I didn't already have a 50MHz 030 in my A1200 I would have put a Vampire in it. Cheaper and faster than an 060 card, has RTG, Ethernet and fast IDE built in, runs cooler...

[Thorham]

Originally Posted by Bruce Abbott

If I didn't already have a 50MHz 030 in my A1200 I would have put a Vampire in it.

If it's a Blizzard then that's nicer anyway.

[OlafSch]

@pandy71 AROS is not Amiga? Wow Now i know that finally ;-) you are clever obviously

Amiga is a hardware and AROS is a OS so both are not identical

What do you want to explain with that?

Don't forget about FPGA Amiga installed base... [/SARCASM]

And what do you want to explain us with that?

We can discuss about AROS 68k if you want, I know it quiet well

there is no general definition of Amiga today that everybody agrees. For some only original A500 or A1200 are Amiga, for others also expanded systems like with V4 or PiStorm or even pure new hardware based on FPGA. You are obviously one of the very traditional ones.

[OlafSch]

Originally Posted by A500

An old guy says something funny in a pub and it is released on Youtube. The video is posted here. 56 forum pages of heated argumentation that shows no signs of ending

the never ending story ... but without a princess

[Promilus]

Originally Posted by Bruce Abbott

All limitations are 'artificial'.

Nope. If logic is designed to handle addresses up to 2MB but for cost reduction you limit it by cutting off one address line so it only handles 1MB that's artificial. No matter how YOU look at it

They didn't need upgrading.

Sure ESD unprotected CIAs didn't ... and neither did aging Paula which couldn't achieve "cd quality" which was all the rave back then. Yay. Just because Bruce said so it DID NOT need an upgraded version.

As I said before, you can't expect the same from a maturing technology. In 1985 the Amiga managed to fill a niche others hadn't fully exploited with an innovative design. By 1992 several companies were doing nothing but designing graphics chips.

Yup and most of those companies are long dead by now. NV was a startup company with no previous experience, small market share and look where they're now. Commodore lacked in innovation and development of new products. So they had a golden goose and let her loose. It's not something you can brush off as "it was obvious Amiga wouldn't be able to keep up with world". Sure it could've. If she was developed by someone who actually tries to stay ahead and keep being innovative. Commodore was not.

The A1000 only 'destroyed' PCs in some areas.

Obviously not in CPU performance overall and certainly not in productivity software. But guess what - sound cards and graphic cards were not really developed to handle better spreadsheet!

That's why I didn't mention it. The C64 lasted longer than the C128, which was never designed to last very long. The C128 was a horrible kludge using the graphics chip from their failed C900 design, and a Z80 jammed into it to do CP/M.

The successor to the C64 would have been the C65. This was quashed for being too much like the Amiga but not as good - one example of Commodore of acting sensibly and us being the poorer for it!

Actually making C128 as actual upgrade (so improved VIC working well with 2MHz main clock and/or VDC with sprites, DMA and better memory management) would've been great. And should they abandon whole Amiga deal C65 might easily overtake "Atari 500" despite design slightly less powerful than original amiga (which also was pimped up by Commodore during implementation). And that's because backward compatibility = great library of existing games. And no pressure to be something else than just cheap home gaming computer.

Launched in 1994, 7 years after the Archimedes.

again, there were multiple models along the way (some with HDD stock and some with ARM3 instead of ARM2).

In 1992 VLB was only available in high-end 486 machines. The Amiga had a 32 bit bus since 1990 with the A3000, and graphics cards that used it started appearing in 1992.

nevertheless it was available. And even before that PC already had graphic cards capable of handling high res modes with highcolor... just not as fast as 1st person shooters requires.

They were expensive for sure, but if price wasn't a concern...

It never is. As long as it was commercial product with support - it will sell. That's why during pandemic ppl were buying top graphic cards even if they were sold at ~2000 or even 3000$. In '92 Cryo released Dune CD and Westwood released Dune II on CD. For that fraction of all PC users which HAD CD-ROM installed in their PCs which at that time required fairly recent sound card as well. And yet they targeted small market for expensive PC rigs... with pretty good effect because iirc both titles did sell well.

OTOH if price was a concern then the A1200 beat the pants off a typical 386SX that cost a lot more.

Well, yeah, and SNES basically beat pants off A1200 despite being rather ascetic by design.

Let's get real here. The Amiga was originally designed to be a home computer like the C64 and Atari 800

Amiga was originally designed as gaming console. And it's first implementation as a computer doesn't really look like "home" computer C64-alike.

The chipset was never designed for high-end stuff and trying to make it so was a waste of time, as Commodore's engineers eventually realized.

I guess marketing team bringing Warhol to the show was in wrong then...

But AGA did give the A1200 better graphics than anything else in its class.

That kind of class was dying off. Ppl which didn't want to spend much and be able to play went to PSX. Ppl which did want to play but also do something else swapped to Mac or PC. A1200 was never nearly as popular as the biggest system seller, A500. Not A1000, not 2000, certainly not 3000. A500. If you sum up all other Amiga models sold they'd hardly match A500 alone. You can have fond memories of A1200 but statistically when someone has fond memories of Amiga it's A500 user.

I have to disagree. Other home computers 'stagnated' just as much, and fans didn't complain because they liked that stability. When the A1200 came out many A500 users argued that their existing machines were good enough.

That's because of stagnation ... there were few games showing of benefits of new platform. And most of the titles didn't look all that better than what A500 did offer. Should there be gradual improvement over time it would've been pretty obvious regular A500 with 512KB CHIP RAM loses to A500 with 1MB chipram and A1200 with 2MB of chipram wins over both of those. But 1MB chipram was ECS standard which en-masse came with A500+ and shortly after with A600. That's 1991/92.

Funny. I have a New Zealand PC magazine from 1994 with an article arguing that buying the latest tech was not a good idea. They suggested buying a 386SX in 1992 and waiting until 486's got cheaper. The example showed significant savings to a business that did that.

Funny. I can find articles saying PC is doomed. I can find articles saying PSX will be sh*t. I can find articles saying Rise of the Robots will be astonishing. And articles saying how all CISC are heading for dumpsters.

If it wasn't for Commodore's efforts there would be no Amiga. They were committed to it until the end. Your lack of appreciation for all the hard work the people at Commodore did in order to deliver wonderful Amiga models unto our hands shows who is being shitty.

From the original Amiga team many members didn't last more than few years. Ask them why they left. Some already explained in interviews. What engineers did want and did achieve is not the same thing as what Commodore did and did not. I'm not talking R&D, I'm talking actions of whole company = decisions of CEO. That's what generally happens when CEO wants firm control but poorly understand the market and the product.

[Rotareneg]

Originally Posted by Promilus

Sure ESD unprotected CIAs didn't ... and neither did aging Paula which couldn't achieve "cd quality" which was all the rave back then. Yay. Just because Bruce said so it DID NOT need an upgraded version.

Don't forget a serial port that appears to have be an afterthought.

[pandy71]

Originally Posted by OlafSch

@pandy71 AROS is not Amiga? Wow Now i know that finally ;-) you are clever obviously

Amiga is a hardware and AROS is a OS so both are not identical

What do you want to explain with that?

Don't forget about FPGA Amiga installed base... [/SARCASM]

And what do you want to explain us with that?

We can discuss about AROS 68k if you want, I know it quiet well

there is no general definition of Amiga today that everybody agrees. For some only original A500 or A1200 are Amiga, for others also expanded systems like with V4 or PiStorm or even pure new hardware based on FPGA. You are obviously one of the very traditional ones.

Olaf, i highly respect you and highly appreciate your work but there is no need to be so tense, relax - i will repeat - AROS is not Amiga - AROS can run on many machines, mostly on modern PC's and PC is not Amiga. AROS is a implementation of the Amiga OS (sort of) - quoting wikipedia:

AROS Research Operating System (AROS, pronounced "AR-OS") is a free and open-source multi media centric implementation of the AmigaOS 3.1 application programming interface (API).

Amiga is Amiga, Amiga OS is Amiga OS, AROS is AROS - that's all.
AROS is not Amiga substitute or replacement.
FPGA Amiga even with some HW extensions (like SAGA) are still Amiga.
And if i can agree that there is no general definition what is Amiga today then i think it is way easier to say what is not Amiga - PC's are not Amiga, Macintosh is not Amiga, smartphones are not Amiga even if you can replace smartphone OS somehow and implement some emulation of Amiga on smartphone.
And it is not about tradition or not - if generic PC can run AROS and after while Mac OS it doesn't mean that this PC is Amiga and/or Macintosh.

[pandy71]

Originally Posted by Bruce Abbott

I have a Vampire V2 in my A600. It's twice as fast as the A3000 I used to have with 50MHz 060, has 4 times the RAM, faster RTG, and is more compatible. And way way cheaper. I tried other accelerator cards in the A600 and they were underwhelming - not the Vampire. It's great for browsing the web and playing those ported games Duke Nuken 3D and Quake. Good for emulators too. I even managed to play Tomb Raider (slowly) under PC Task!

The latest V4 cards are a bit more expensive, but even more powerful and have more features. If I didn't already have a 50MHz 030 in my A1200 I would have put a Vampire in it. Cheaper and faster than an 060 card, has RTG, Ethernet and fast IDE built in, runs cooler...

Ok, don't get me wrong but fact that 060 accelerators are insanely priced doesn't mean that V4 has good price/performance ratio.
My point was to not criticize Apollo team marketing strategy but only to say that even after spending like 1000€ on extending Amiga you are still unable to compete with 1000€ PC.

[pandy71]

Originally Posted by Promilus

Nope. If logic is designed to handle addresses up to 2MB but for cost reduction you limit it by cutting off one address line so it only handles 1MB that's artificial. No matter how YOU look at it

lol - tell this to Motorola using full 32 bit address but exposing only 20 or 24 bits externally (depends on CPU). This is normal practice... also nowadays, sometimes chips are artificially crippled - nowadays are even more strange practices as for example dark silicon.

Originally Posted by Promilus

Sure ESD unprotected CIAs didn't ... and neither did aging Paula which couldn't achieve "cd quality" which was all the rave back then. Yay. Just because Bruce said so it DID NOT need an upgraded version.

Of course not upgrading Paula was severe Amiga limitation - Paula definitely should be upgraded at least in AGA.

And all Amiga custom IC's are ESD sensitive - CIA's are not exception and probably ESD is not primary CIA's concern (as there is some level of the ESD protection implemented on PCB by Commodore) but lack of latch-up protection.

[OlafSch]

@pandy71

I seem to have missed something... who claimed AROS is Amiga? AROS is a replacement for AmigaOS, not more not less

AROS running on PC is of course not Amiga, it is still a PC with a obscure OS, not Amiga