Should the A1200 have used a 28Mhz 68000 processor instead?

[jimbob]

Originally Posted by eXeler0

Welll.. you know.. it's never too late to write an e-mail and remind him to go f-himself ;-)

http://stoneridgepartners.biz/contact.htm

"His prior experience includes serving as the President of Commodore International, where he accomplished a major operational turnaround."

I guess that is one way of putting it.

[eXeler0]

Originally Posted by jimbob

"His prior experience includes serving as the President of Commodore International, where he accomplished a major operational turnaround."

I guess that is one way of putting it.

Just as clueless now as he was in his time in charge at Commodore...
I wonder if he is proud of his achievements... All the more reason for Amigans to remind him that he sucked and ruined something that was revolutionary... 😆

[frank_b]

Originally Posted by gulliver

CBM released the A1200 at 14MHZ because they got cheaper pricing on those 68020 chips. A 28mhz would have been more expensive.

Instead of going to 28MHz it would have made more sense to add 4MB of fast ram to it. With that enhancement alone, The A1200 would have doubled its speed an enabled the use of more "serious" applications.

The chip RAM bus speed is only 7 mhz. It wouldn't have made sense to use a faster chip without adding fast RAM. The only benefit would have been ROM access.

[Samurai_Crow]

The easy choice would have been a clock quadrupled blitter that took advantage of fast page fetching. A fast copper could have made sense once the fast page buffer circuit existed. Running the fast bus at a higher clock would have made sense only after the chip bus bandwidth was maximized.

[lordofchaos]

Originally Posted by Samurai_Crow

The easy choice would have been a clock quadrupled blitter that took advantage of fast page fetching. A fast copper could have made sense once the fast page buffer circuit existed. Running the fast bus at a higher clock would have made sense only after the chip bus bandwidth was maximized.

Good idea! And a smart one, but history shows us that smart isn't always the right choice, PC solution was always raw power, not efficient or clever inventions, still to this day it seems

[delshay]

I have a few 68EC025 version of this chip (new) Freescale.

[eXeler0]

Originally Posted by lordofchaos

Good idea! And a smart one, but history shows us that smart isn't always the right choice, PC solution was always raw power, not efficient or clever inventions, still to this day it seems

Well it's a little late for good ideas now, isn't it. ;-) Also it usually easier to point out the mistakes afterwards than propose the clever solutions at the beginning.

(It will be interesting to see how the SAGA/Apollo-core guys will solve an improved blitter/copper situation now that history has provided us with the examples.)

Regarding the PC philosophy.. We need raw power to compensate for the ever increasing "bloat"...
Allow me to whine a bit:
<START whine>
When I started out with 3d on Amiga 20 years ago, I used "Imagine" which consisted of a single 1MB file. It had editors, modeling tools, ray tracing engine etc..
Now I work with Autodesk 3ds Max and when I install it, I get Gigabytes of shit around the main application.. stuff that I don't even need and installation creates thousands of files and folders... Each new release is so full of bugs that it's almost unusable until the first service pack arrives. Then m installing a service pack onto an ssd takes forever and the release notes take more space than the entire "Imagine" application did..
Code Base is so complex (and messy) that there is no room for clever and efficient inventions. I get the feeling the devs are thrilled just to see this dinosaur compile to something that even runs at all.. 😡
<END whine>

[idrougge]

Originally Posted by Samurai_Crow

The easy choice would have been a clock quadrupled blitter that took advantage of fast page fetching. A fast copper could have made sense once the fast page buffer circuit existed. Running the fast bus at a higher clock would have made sense only after the chip bus bandwidth was maximized.

I think that's what AA+ was supposed to do, running off a 57 MHz clock with 140 ns bus cycle. That new chipset was poised to be introduced quite quickly after AGA, but economic realities came between.

[teppic]

The 68020 was hugely important as it was fully 32bit, which was holding back the 68000 a lot. Fast 68000s with a 16bit data bus were a bit pointless beyond the stock 7MHz - you wouldn't get much of a performance increase. On the A1200 I think chip RAM had 24bit access to the custom chips, but it did have 32bit access to the CPU.

The problem really was the massive delay between the A500 and the A1200. Commodore went from 1985 to 1991 with no real updates -- by 1991 they were releasing the A3000 at ridiculous prices with a chipset that was several years out of date, and the A500 still on 1985's tech.

AGA could well saved the day if it'd come out in 1990 and Commodore had never launched the A600.

[TjLaZer]

The 1200 with the 020 was fine even with only 2 MB of chip memory as that was the foundation for upgrading the machine faster. Simply adding a 32-bit memory card to the machine helped a lot along with faster accelerator and RAM cards as we all know. I see nothing wrong with what they did, it was an entry-level home computer.

[TjLaZer]

I would argue it was the A500+ that should have had a faster 68000

[Amigajay]

The A500 selling well into the early 90s had Commodore in a quandry, A500 sales were increasing every year so they had to decide to break this cycle and risk losing potential sales.
The A500+ was an awful machine to put out there, esp as the A600 was planned anyway to put production costs down, they should have waited for that (im guessing it was commodores way of trying to get rid of the excess A500 cases and motherboards) but no way should they have a faster cpu in it, it would have killed sales and alienated the millions of A500 owners in one swoop.

The A1200 was fine as a budget computer and competed pretty well against the consoles, and would have served and lasted longer obviously with Commodore alongside, and yes it would have been better coming out in 1990, but the machine in itself was pretty sound for the price, you cant in hindsight want to add fast ram and faster cpu and still sell it for the same price in 1992s prices.

[eXeler0]

Originally Posted by Amigajay

The A500 selling well into the early 90s had Commodore in a quandry, A500 sales were increasing every year so they had to decide to break this cycle and risk losing potential sales.
The A500+ was an awful machine to put out there, esp as the A600 was planned anyway to put production costs down, they should have waited for that (im guessing it was commodores way of trying to get rid of the excess A500 cases and motherboards) but no way should they have a faster cpu in it, it would have killed sales and alienated the millions of A500 owners in one swoop.

The A1200 was fine as a budget computer and competed pretty well against the consoles, and would have served and lasted longer obviously with Commodore alongside, and yes it would have been better coming out in 1990, but the machine in itself was pretty sound for the price, you cant in hindsight want to add fast ram and faster cpu and still sell it for the same price in 1992s prices.

As much as I love my 1200 these days.. I actually *hated* the specs when it came out. After the A4000 I was expecting something it the 25MHz range (020 or *maybe* 030 obviously... not 040) and a memory configuration that could utilise that power... I wouldn't minded £50-100 higher price for a higher clocked CPU and say 2MB fastram on the mobo..
As it was... a couple of months with a vanilla A1200 I got the Blizzard 1220 with 4MB Fastram and at that point I felt I finally had a usable Amiga that I could use for other stuff than games ;-) (and of course a bunch of polygon games became playable)...
In the end.. I suppose Commodore didn't wanna risk missing the sweet spot (price wise) and for that I can't blame them. But looking at it objectively.. they moved forward far to slow.. the 90's weren't as forgiving as the 80's. The development pace picked up and a company like Commodore never had a good chance in the new fast moving world...

[Amigajay]

Originally Posted by eXeler0

As much as I love my 1200 these days.. I actually *hated* the specs when it came out. After the A4000 I was expecting something it the 25MHz range (020 or *maybe* 030 obviously... not 040) and a memory configuration that could utilise that power... I wouldn't minded £50-100 higher price for a higher clocked CPU and say 2MB fastram on the mobo..
As it was... a couple of months with a vanilla A1200 I got the Blizzard 1220 with 4MB Fastram and at that point I felt I finally had a usable Amiga that I could use for other stuff than games ;-) (and of course a bunch of polygon games became playable)...
In the end.. I suppose Commodore didn't wanna risk missing the sweet spot (price wise) and for that I can't blame them. But looking at it objectively.. they moved forward far to slow.. the 90's weren't as forgiving as the 80's. The development pace picked up and a company like Commodore never had a good chance in the new fast moving world...

Yeah I'm not saying it was the best at the time, for the price the specs were ok, but the next Amiga lived or died on its custom chips, which sadly Commodore didn't do enough with, getting rid of the original team was a massive mistake and they thought any old engineer team could come up with the next generation...in hindsight we know this wasn't to be the case. That's something at least Atari got right, getting decent engineers/buying designs i.e flare technology, they were just shit at marketing, Commodore (at least in Europe) were the other way around.

[Lord Aga]

Originally Posted by TjLaZer

I would argue it was the A500+ that should have had a faster 68000

A600 also. With a 14MHz 68000 it would be a nice little beast

[Mrs Beanbag]

the ROM changes were a bigger problem for compatibility, i think. which made me wonder lately if it's possible to make a version of 3.x ROMs built for compatibility with dodgy 1.3 software, just by making sure all the functions start at the same absolute addresses and whatnot.

[teppic]

The 68EC020 was mostly fine. It had the same 24bit address bus as the 68000, so badly written code would usually work without an issue. By the 90s the 68000 was extremely out of date though, even with a high clock rate it lacked any instruction cache, and its 16bit data bus limited its performance as memory increased. I very much doubt it could keep up well with the demands of the higher performing A1200 chipset.

If the A1200 had come out just a year or so earlier, bypassing the A500+ and the A600 completely, Commodore could have launched a new graphics chipset with a new ROM, and people would have seen it as a major upgrade, ignoring small incompatibilities with older software.

edit:

I was just looking at wikipedia, and apparently AGA had a full 32bit data bus access and a doubled clock rate. That would also explain why a 68020 was matched with it (also having a 32bit data bus for the RAM).

[eXeler0]

Originally Posted by teppic

I was just looking at wikipedia, and apparently AGA had a full 32bit data bus access and a doubled clock rate. That would also explain why a 68020 was matched with it (also having a 32bit data bus for the RAM).

The interesting(?) question is why Atari chose a 16MHz 030 while using a 16-bit data bus and 24 bit adress bus. Couldn't they have saved a penny by using a 020 instead. Cheap ass Commodore aparently thought it was worth chosing a 020. (I have no idea what the prices were for 020 and 030 chips respectively at the time).

Edit: I'm talking about Atari Falcon which is contemporary with the A1200.

[frank_b]

Originally Posted by eXeler0

The interesting(?) question is why Atari chose a 16MHz 030 while using a 16-bit data bus and 24 bit adress bus. Couldn't they have saved a penny by using a 020 instead. Cheap ass Commodore aparently thought it was worth chosing a 020. (I have no idea what the prices were for 020 and 030 chips respectively at the time).

Edit: I'm talking about Atari Falcon which is contemporary with the A1200.

I suspect the Falcon was originally supposed to have a 68000 in it. Remember though that the Falcon has a 16 mhz 16 bit bus. Without fast RAM the Falcon is faster at 16 bit memory accesses than a 1200. It's slightly faster than a stock 1200 on 32 bit accesses too. It's a moot point anyway, you can add fast ram and better CPUs to both easily enough. Having a full 030 on the Falcon gave them access to memory protection in Mint/MultiTOS.

[eXeler0]

Originally Posted by frank_b

I suspect the Falcon was originally supposed to have a 68000 in it. Remember though that the Falcon has a 16 mhz 16 bit bus. Without fast RAM the Falcon is faster at 16 bit memory accesses than a 1200. It's slightly faster than a stock 1200 on 32 bit accesses too. It's a moot point anyway, you can add fast ram and better CPUs to both easily enough. Having a full 030 on the Falcon gave them access to memory protection in Mint/MultiTOS.

Interesting theory. Atari must have made some last minute "panicy" design decisions when they found out about Commodore was working on the A1200.