"Looking forward" to the Amiga 1200...

[Daishi]

I was lucky enough to get an A1200 before official release (I knew a guy) by 2 weeks. When I got it home my excitement took a rather bad downturn as my friends A500 with an A5000 accelerator was faster. it's not that the A1200 was a bad computer it just wasn't a great one out of the box. When I bought mine I was disappointed compared to my friends upgraded A500 (68020 @16 Mhz but on a 16-bit bus) the A1200 with a 68020 @ 14Mhz (32-Bit bus) didn't feel any faster. I ordered a memory expansion card and that made a big speed difference. That got me on the upgrade path Next a 68020 @28 Mhz followed by a 68030 @ 40Mhz then 50Mhz then a 68040 @25Mhz and finally a 68060 it was a beast for applications but games rarely took advantage of that power.
A quick story - I worked in the IT dept at a school in London we had BBC B, Archimedes, Macs and PC's. I would bring my Amiga in (68030 @50 at the time) and run emulators of all the systems on the Amiga (except Arch that machine was a beast in its own right) and while the PC emulation was slow it was usable for what I was doing. However, the other emulators were as fast if not faster than the systems being emulated and that included the Mac. The IT director was a huge Mac fan and one day saw me using my Amiga and asked what I was doing, so I explained that it was easier for me to use the emulators to create work across the systems. I then showed her running Shapeshifter and doing all the same work we used them for. Then without quitting I jumped to the BBC emulator and ran a program I had done, while the program was running I jumped to the PC emulator and ran a game (just some adventure game we used at school) and just for giggles I jumped back to the Amiga loaded a MOD to play, formatted a floppy and while it was doing all that, jumped back to the Mac and proceeded to use Pagemaker and an Art package (I think it was Photoshop) the look on her face was priceless she didn't know what to say, she just flubbered someting and left.
The Amiga was (is) a special system from a time when other systems just couldn't do those things.

[Tigerskunk]

Originally Posted by NorthWay

31 years later I got my own A1000. It still has killer looks but my mind insisted it was larger than it turned out to be.
IMO it could have been scaled up a bit in both x(5-10cm)/y(?)/z(2-3cm) and retained the look, but also having space for Zorro cards.

I got mine last summer, and was surprised it was so big.

Remembered it being a bit smaller than it was.

Haven't seen one since 1986, though...

[Hedeon]

Would I spoil it more if I said the address bus was 24 bit of the 68EC020?

[Gorf]

Originally Posted by Hedeon

Would I spoil it more if I said the address bus was 24 bit of the 68EC020?

No that was already mentioned here.
And while a maximum of 16 MByte was fine in 1992, no FastRAM and only 14MHz were not.

The Atari Falcon from the same year had a 68030. Clocked with 16 MHz, but as rumor goes Motorola could only deliver 33MHz versions and so the CPU clock in almost every Falcon can be doubled...
But Atari build in an equally stupid showstopper: the data bus is only 16 bit wide!

[Foebane]

I remember getting my Amiga 1200 for my birthday. My father came up to tell me he'd gotten me a Scalextric game, but it was the Amiga 1200 pack with the Nigel Mansell game in it. He'd also put a cushion on top of the box, and when I took it off, there was a big bold sticker stating "85Mb Hard Drive" and I was quite delighted by that. It was my first hard drive, and I loved the convenience of it.

[trixster]

Originally Posted by Gorf

The Atari Falcon from the same year had a 68030. Clocked with 16 MHz, but as rumor goes Motorola could only deliver 33MHz versions and so the CPU clock in almost every Falcon can be doubled...

I think that is probably true - the BlowUP accelerators 'overclock' the stock CPU to 32/36/40Mhz; if the CPU were only rated to 16Mhz then I doubt such an "accelerator" would have existed.

[d4rk3lf]

Question:
Let's imagine (very hypothetically ) Commodore had the best moves they can have. They release (better version) A1200 few years earlier, then AAA chipset with new Amiga... etc... they earn a tons of money and become leader on the market.

But Motorola stops 68xxx processor production after 060 (as they did). What could Commodore do then? What processors you think they'd choose? How would people that had Amiga feel about that?

[idrougge]

Either PA-RISC, since they already had some cooperation with HP and there were talks about moving to PA-RISC back then, or the PowerPC, since it was competitive, Apple had already spearheaded the transition and they already were a big Motorola customer.

[TEG]

Do you remember if there was a version of the 1200 proposed with the HD included ? Because for me it was the big news for the Amiga, the possibility to have an integrated HD. All PC were running with an HD at this time and part of their power and attractiveness came from this. It was life changing at the time and if I remember well the 2"1/2 HD needed for the 1200 was very much expensive than the 3"1/2 needed for the PC. Anyway, I did not remember a 1200 HD ready and a nice WB already configured and bundled with applications, being pushed by C=

[d4rk3lf]

@idrougge
Thanks

@TEG
I don't think so, but I might be wrong.
Imagine they had made A1200 with 030, 4-8 MB Ram, and a HD 40-80MB. That would be awesome machine (I think even on 1992).

[TEG]

@TEG
I don't think so, but I might be wrong.
Imagine they had made A1200 with 030, 4-8 MB Ram, and a HD 40-80MB. That would be awesome machine (I think even on 1992). [/QUOTE]

I wonder of the price difference (when bought in mass) between 68020 and 68030 in 1992. However, to ponder, we have to remember that the Falcon did not sell.

[idrougge]

There were several A1200 bundles with hard disk, as evidenced by some A1200s with special "A1200/HD80" etc badges.

[Gorf]

Originally Posted by TEG

@TEG
I don't think so, but I might be wrong.
Imagine they had made A1200 with 030, 4-8 MB Ram, and a HD 40-80MB. That would be awesome machine (I think even on 1992).

I wonder of the price difference (when bought in mass) between 68020 and 68030 in 1992. However, to ponder, we have to remember that the Falcon did not sell.

The Falcon was considered a step backwards (16Bit) to the full 32Bit TT.

At one side it was total overkill: DSP (plus DSP-port), IDE and SCSI, Expansion Port, 2 serial ports plus extra MIDI in and out, 2 digital mouse/joystick plus 2 analog ports, Local Talk port...

And on the other side it stupidly used a 16Bit data-bus for 32bit CPU and DSP!

That combination made it expensive an slow at the same time.

For the 1200:
The price difference to a 68ec030 could have not been more than 50$ probably less.
And "Budgie" in the 1200 is already capable of managing 32Bit FastRAM - just add some cheap (50 cent) SIMM-sockes and start with 1 MB Chip and 1 MB Fast (both upgradable with SIMMs)

[sneeker]

I wanted a 1200, I had a 500 and a friend got a 1200 for Chrismas, Played on it quite a bit and I was due to get one for I think Christmas 93 with a 120mb hdd, think is was about £520ish at the time...

Then I looked at pc's and found I could get a 386/40 with monitor for about the same price. That gained a soundcard and some ram over the next year and Then got upgraded to a Dx2/66 the following Christmas.

[Gorf]

Originally Posted by d4rk3lf

Question:
Let's imagine (very hypothetically ) Commodore had the best moves they can have. They release (better version) A1200 few years earlier, then AAA chipset with new Amiga... etc... they earn a tons of money and become leader on the market.

But Motorola stops 68xxx processor production after 060 (as they did). What could Commodore do then? What processors you think they'd choose? How would people that had Amiga feel about that?

PPC and PA-RISC were already mentioned. MIPS like in the PS1 and PS2 would also been a reasonable choice.

On the other hand: Motorala was already developing the 68060B with some optimizations (pipelined FPU) and higher clock rates. They surly would have build it for the "marked leader"
After that maybe a deal with Transmeta to emulate the 68K ISA in addition to x86.
(I was in email-contact with Transmeta back than and they wrote me, that 68K was in consideration earlier and absolutely possible, but the market not big enough.)

[Daedalus]

Originally Posted by Gorf

The price difference to a 68ec030 could have not been more than 50$ probably less.

The A1200 was already expensive - adding such a significant extra cost to the machine would have meant far fewer sales. If you think that Commodore were saving money by not including the Workbench install disk with a computer that could be fitted with a hard drive, an extra $50 is a big deal. Don't forget as well the cost of the additional development to fit the 030 on the board.

And "Budgie" in the 1200 is already capable of managing 32Bit FastRAM - just add some cheap (50 cent) SIMM-sockes and start with 1 MB Chip and 1 MB Fast (both upgradable with SIMMs)

You seriously underestimate the additional costs involved in a change like that. More assembly steps, more parts, more testing, more expensive memory, more warranty repairs... Plus, most people who bought such Amigas were stuck in the 8-bit mentality of not upgrading and just playing games, so 1MB chip RAM would've remained the standard, further limiting those few games that did bother to target AGA machines specifically.

[Amigajay]

Originally Posted by d4rk3lf

@idrougge
Thanks

@TEG
I don't think so, but I might be wrong.
Imagine they had made A1200 with 030, 4-8 MB Ram, and a HD 40-80MB. That would be awesome machine (I think even on 1992).

No one would have bought it as it was getting into low end PC prices at that time, dont forget Atari were good at getting tech out at decent prices and even then the Falcon with only 1mb ram cost £499 or £899 with 4mb and 65mb hdd, so i imagine a similar spec A1200 would be £1000, no way that would ever sell with only a 030 and AGA specs.

[Gorf]

Originally Posted by Daedalus

The A1200 was already expensive - adding such a significant extra cost to the machine would have meant far fewer sales. If you think that Commodore were saving money by not including the Workbench install disk with a computer that could be fitted with a hard drive, an extra $50 is a big deal. Don't forget as well the cost of the additional development to fit the 030 on the board.


You seriously underestimate the additional costs involved in a change like that. More assembly steps, more parts, more testing, more expensive memory, more warranty repairs... Plus, most people who bought such Amigas were stuck in the 8-bit mentality of not upgrading and just playing games, so 1MB chip RAM would've remained the standard, further limiting those few games that did bother to target AGA machines specifically.

I do not think, that the 1200 was in any way expensive. A 386-PC, the Falcon or any Mac were more expensive in 1992.
Commodore knew already very well how to fit a 68030 into a machine. So there a no higher development costs compared to the 68020.
There would be no justification for a higher price tag, than the plain <50$ the chip costs more.

There are not really more parts or more expensive memory involved. The testing and so on would have been exactly the same as they were.

Almost every A500 was upgraded to 1 mb.
The same users we are talking about had suddenly no problem upgrading their PCs after leaving the Amiga. A better Amiga would have stopped many from leaving.

[Daedalus]

Originally Posted by Gorf

I do not think, that the 1200 was in any way expensive. A 386-PC, the Falcon or any Mac were more expensive in 1992.

And all were more capable, at least from a hardware perspective. For what it was - a home computer without a hard drive, primarily intended for games use - it was expensive (and even moreso during Escom's tenure, when its price increased even though it was even more out of date). Increasing the price much more would have a serious impact on sales as it was already on the expensive end for parents buying them for their kids' Christmas presents.

Commodore knew already very well how to fit a 68030 into a machine.

Indeed, the design principles aren't any problem, it's the physical fitting of the chip on the board that's the problem as there's not much space in that area to spare. True, if it had been designed from scratch as an '030 machine that would be already factored in, but the aim was to produce a cheap machine, so not having an '020 as at least an option would be against that brief.

So there a no higher development costs compared to the 68020.

What about the extra prototyping costs?

There would be no justification for a higher price tag, than the plain <50$ the chip costs more.

You don't work in the manufacturing industry, do you? Anyway, as I said, it was already an expensive computer - too expensive for me and many others to get at Christmas for example. An extra $50 would have put it out of reach of even more customers.

There are not really more parts or more expensive memory involved. The testing and so on would have been exactly the same as they were.

You're proving my point that you don't really have any idea of the differences involved.

Extra SIMM sockets: That's extra through hole soldering, extra testing during manufacture to ensure it works with other modules than the ones it shipped with, extra assembly time since the SIMMs have to be fitted, all on top of the cost of the sockets themselves. There will also be extra manufacturing failures requiring higher numbers of reworks or scraps.

SIMMs themselves: No matter how you cut it, buying RAM chips in tubes will always be cheaper than buying those same chips already soldered to extra PCBs and QC tested. Soldering to the board will always be cheaper - look at bargain laptops these days with 4GB of RAM soldered to the mainboard instead of fitting a cheap SO-DIMM socket.

Warranty issues: More parts and more connectors mean more failures. It's one of the reasons chips (other than Kickstart) aren't socketed on the motherboard either.

Almost every A500 was upgraded to 1 mb.
The same users we are talking about had suddenly no problem upgrading their PCs after leaving the Amiga. A better Amiga would have stopped many from leaving.

And why is it that those same users didn't buy any of the many readily-available upgrade cards for the A1200? It could be upgraded similarly to their new PCs - faster CPUs, more RAM, hard drives. Yet very few did. A 1MB A500 was a necessity for playing many games; an '030, 6MB A1200 was not.

[Gorf]

Originally Posted by Daedalus

And all were more capable, at least from a hardware perspective. For what it was - a home computer without a hard drive, primarily intended for games use - it was expensive (and even moreso during Escom's tenure, when its price increased even though it was even more out of date). Increasing the price much more would have a serious impact on sales as it was already on the expensive end for parents buying them for their kids' Christmas presents.

not much more but around 50$.

Indeed, the design principles aren't any problem, it's the physical fitting of the chip on the board that's the problem as there's not much space in that area to spare. True, if it had been designed from scratch as an '030 machine that would be already factored in, but the aim was to produce a cheap machine, so not having an '020 as at least an option would be against that brief.

That's what I am talking about! Planing it right from the start.
No: a 68020@14Mhz was not an option for an machine in 1992.
(as mentioned earlier you would by than find better CPUs in printers and washing machines.)

What about the extra prototyping costs?

no additional prototyping.

You don't work in the manufacturing industry, do you? Anyway, as I said, it was already an expensive computer - too expensive for me and many others to get at Christmas for example. An extra $50 would have put it out of reach of even more customers.

the main problem in 1992 was, that they could not deliver enough units, due to too few chipsets ordered. I am sure a the demand for a 68030 version would have been ever greater, even with 50$ more on the price!

The same people you are talking about had no problem to spend even more money on a PC, when no adequate Amiga was available.

A more powerful A1200 at a little bit higher price would not have been too expensive but just right. People were manly disappointed by the low specs and not by the price.

(Of course a low-cost desktop "A1000 NG" would have been an even better idea...)

You're proving my point that you don't really have any idea of the differences involved.

Extra SIMM sockets: That's extra through hole soldering, extra testing during manufacture to ensure it works with other modules than the ones it shipped with, extra assembly time since the SIMMs have to be fitted, all on top of the cost of the sockets themselves. There will also be extra manufacturing failures requiring higher numbers of reworks or scraps.

you can solder the first megabites in directly to avoid this.
There are SMD-SIMM-sockets, so no through hole soldering.
You could save the clock-port instead.

SIMMs themselves: No matter how you cut it, buying RAM chips in tubes will always be cheaper than buying those same chips already soldered to extra PCBs and QC tested. Soldering to the board will always be cheaper - look at bargain laptops these days with 4GB of RAM soldered to the mainboard instead of fitting a cheap SO-DIMM socket.

I never meant to put the default RAM on SIMMs, but wrote about an easy way to upgrade, for users or dealers.

Warranty issues: More parts and more connectors mean more failures. It's one of the reasons chips (other than Kickstart) aren't socketed on the motherboard either.

see above

And why is it that those same users didn't buy any of the many readily-available upgrade cards for the A1200? It could be upgraded similarly to their new PCs - faster CPUs, more RAM, hard drives. Yet very few did. A 1MB A500 was a necessity for playing many games; an '030, 6MB A1200 was not.

I do not know about the ratio of people upgrading their A1200. Do you?
It was sold in much fewer numbers than the A500, in part due the reasons mentioned above. So it might give the impression only few people upgraded due to overall fewer sales.

Also it was much cheaper für PC users to upgrade their RAM, so they did more often. Thats the reason I would have preferred a SIMM socket on the board.

So to summarize your point:
there was nothing Commodore could have done to generate more sales?