Was anyone else disappointed with the A1200?

[Promilus]

@Karlos - Amiga chipset was elegant. But by after a while it was exactly what backfired. Trying to add much more new features wouldn't leave backward compatibility intact. And so the only other choice was to slightly improve it and fight with price rather than performance. It did however backfired as well as by that time Amiga was reduced to budget gaming platform (ofc there were some niches for big Amigas but it soon dried out as well and obviously amiga chipset wasn't of much use later on that's why things like DraCo appeared). So... why be interested in something considered worldwide as gaming console with computer keyboard? And not that impressive when paired up against next gen consoles? As I already said for me PC gaming boom was second half of 90s. But it would be false to assume there was a shortage of games for PC platform before that. And I believe - worldwide - PC used for gaming still was bigger market than whole Amiga. So to stay in business Commodore would have to bring up something really impressive. While A1200 was adequate at the moment of it's premiere it didn't leave reviewers speechless. And so it was unable to secure ample userbase and stable flow of cash. Of course there were PC related companies which died out around the same time. Sure. But those were actually competing against each other on the same market. Amiga had it's own market which wasn't really growing anymore and A1200 didn't turn the tide. Hence the disappointment.

[sokolovic]

Promilus, I don't know the sales figure of the A1200 but it seems to me that it solds quite well in its short lifespan considering that this is probably the most common Amiga after the A500 and that it was quite well supported by the developpers (compare it to the Falcon for instance).
So all things considered, the A1200 was a good move from Commodore. It was killed by Commodore bankrupcy but I'm not sure that it was responsible for it.

[Promilus]

but it seems to me that it solds quite well in its short lifespan

introduced in late 92, sold up to 96 probably (as in 97 Escom went down as well). I'm kinda lost with the "short lifespan". 3DO wasn't supported even that long and sold out better despite hefty price tag.

but I'm not sure that it was responsible for it

It was responsible in only one way - it failed to be good enough to keep them running.

[Cyprian]

Originally Posted by Bruce Abbott

I'm sick of hearing about how Commodore was so mismanaged they never managed to do anything. The record proves otherwise. Not only did they have a clear path for the Amiga, they did try to gain new markets and they were constantly bringing out new stuff.

1985 - A1000
1986 - Sidecar
1987 - A500, A2000, A2090, A2052, A2058, A501
1988 - A2620
1989 - A2630, A590, A560
1990 - A3000, A2060, A2065, A2232
1991 - A500+, A501+, CDTV, A2091, A2410, A2320
1992 - A600, A1200, A4000, A570
1993 - CD32

Just to summarize:
1984 - A1000:
- CPU 8MHz
- Bus bandwidth: 3.5MB/s for the CPU, 3.5MHz (1.7MHz for the CPU) 16bit memory bus
- 8bit audio
1994 - A4000
- CPU 25MHz
- Bus bandwidth: 7MB/s for the CPU, 3.5MHz (1.7MHz for the CPU) 32bit memory bus
- 8bit audio

In a meantime they produced tons of the same hardware in a different cases (desktop, .


Look at the Sony:
1994 - PS1:
- CPU 33MHz,
- Bus bandwidth: 132MB/s
2000 - PS2:
- CPU 299MHz,
- Bus bandwidth: 3.2GB/s (1.2GB/s for the video)
2006 - PC3:
- CPU 3.2GHz,
- Bus bandwidth: 25GB/s



If Sony followed the path of the Commodore, in the year 2000 we would have tons of PS1 hardware in a different fancy cases with different stickers and PS2 would have spec: CPU 99MHz with bus bandwidth: 260MB/s


Sounds like devolution

And to be honest, similar story was in my Atari world (besides the Jaguar, which was technologically really advanced - 106MB/s video bus bandwidth in 1992)

[Thorham]

Originally Posted by Bruce Abbott

I'm usually pretty careful with my prefixes, but sometimes I forget. 2MB is a reasonable amount of RAM on an Amiga. On a PC it's nothing. You need 1000 times that to do anything useful on a modern PC. truly mind-boggling, yet most people don't think about it.

It's madness. My new peecee has 32GB of RAM and 12GB of RAM on the video card What's even crazier is how affordable it has become.

[Gorf]

Originally Posted by Cyprian

Just to summarize:
1984 - A1000:
- CPU 8MHz

Unlike the ST we only got 7Mhz...

- Bus bandwidth: 3.5MB/s for the CPU, 3.5MHz (1.7MHz for the CPU) 16bit memory bus
- 8bit audio
1994 - A4000
- CPU 25MHz
- Bus bandwidth: 7MB/s for the CPU, 3.5MHz (1.7MHz for the CPU) 32bit memory bus

To be fair: the A4000 came with FastRAM, which gives you fantastic 11.5MB/s read, 6.1MB/s write

And to be honest, similar story was in my Atari world (besides the Jaguar, which was technologically really advanced - 106MB/s video bus bandwidth in 1992)

AFAIK the Jaguar was not designed in-house ....
(Maybe Commodore should have hired some external developers to help them...)

[TEG]

Originally Posted by Cyprian

Just to summarize:
1984 - A1000:
- CPU 8MHz
- Bus bandwidth: 3.5MB/s for the CPU, 3.5MHz (1.7MHz for the CPU) 16bit memory bus
- 8bit audio
1994 - A4000
- CPU 25MHz
- Bus bandwidth: 7MB/s for the CPU, 3.5MHz (1.7MHz for the CPU) 32bit memory bus
- 8bit audio

Do you have a link for this 3.5MHz A4000 bus speed ? On the Wikipedia page it's indicated 25MHz :

[Gorf]

Originally Posted by TEG

Do you have a link for this 3.5MHz A4000 bus speed ?

That is ChipRAM timings ...

(but as mentioned above: the FastRAM access in the original A4000 with A3640 board are very underwhelming - 11.5MB/s read, 6.1MB/s write. This configuration was a total waste of an otherwise nice CPU)

[Weasel Fierce]

Originally Posted by TCD

Here's a retro YT video for a sub-par 1996 PC: [ Show youtube player ]
It got a fair amount of views so far being only a day old.

Nostalgia has and will be subjective and as Thomas mentioned there will be people who think the first iPhone is the most important technological invention. Right now the 80s/90s nostalgia is where it is at, but there's a huge crowd of PS1/N64 and beyond folks around. I recommend Tech Tangents on YT for things that are even older like 70s machines or calculators from the early 1900s. Of course all machines that have been will stay relevant to some, but the focus or peak will shift for sure.

LGR does a lot of retro PC stuff as well, which is worth checking out. He doesnt seem to do the "Tech Tales" a lot any more but they are really fun dives into the history of various companies.

[sokolovic]

Originally Posted by Promilus

introduced in late 92, sold up to 96 probably (as in 97 Escom went down as well). I'm kinda lost with the "short lifespan". 3DO wasn't supported even that long and sold out better despite hefty price tag.

It was responsible in only one way - it failed to be good enough to keep them running.

Mmm. I think this prove enough how non objective you are.
No,the A1200 wasn't sold UP to 1996 in every corner stores nor it was supported by big and powerful houses like Electronic Arts, Panasonic, Sanyo or LG after the Commodore demise.

I'd even say that the fact there was a company thinking it was a good Idea trying to sell an A1200 in 1995 with 1993 prices prove that the A1200 wasn't this bad actually. And Escom went down because of the PC Market.

Real A1200 lifespan was from late 1992 to mid 1994. After Commodore bankrupcy it lived by inertia.

[TEG]

Originally Posted by Gorf

That is ChipRAM timings ...

OK, I got it : 7MHz / 2 = 3.5 MHz (half timeslots for the chipset and half for the cpu in Lowres).

But now I have a problem with the 1.7 Mhz affirmation. The 68000 needed 4 cycles to fetch the memory. So 3.5 Mhz / 4 = 0.875 Mhz no ?

[Promilus]

Originally Posted by sokolovic

Mmm. I think this prove enough how non objective you are.
No,the A1200 wasn't sold UP to 1996 in every corner stores nor it was supported by big and powerful houses like Electronic Arts, Panasonic, Sanyo or LG after the Commodore demise.

I'd even say that the fact there was a company thinking it was a good Idea trying to sell an A1200 in 1995 with 1993 prices prove that the A1200 wasn't this bad actually. And Escom went down because of the PC Market.

Real A1200 lifespan was from late 1992 to mid 1994. After Commodore bankrupcy it lived by inertia.

If you can prove that A1200 wasn't produced and sold after 94 then I'll admit I was wrong. If you can't prove that... who is the one lacking objectivity here?
Lifespan of hw is usually defined as time between it's debut up to the end of production and/or official support. Not my fault your definition is different.

[Gorf]

Originally Posted by TEG

OK, I got it : 7MHz / 2 = 3.5 MHz (half timeslots for the chipset and half for the cpu in Lowres).

But now I have a problem with the 1.7 Mhz affirmation. The 68000 needed 4 cycles to fetch the memory. So 3.5 Mhz / 4 = 0.875 Mhz no ?

The 3.5 MHz or roughly 280ns are the chipset timeslots in the first chipset, that never changed in later models.
In the A1000, A500 and A2000 the 68000 CPU needs indeed four cycles anyways. That means if the CPU is clocked twice s fast (7MHz) as the Chipset (3.5MHz) it will need access every second Chipset cycle - and that is exactly what happens and why the first three Amigas do not really need FastRAM, as long as you restrict the Chipset to every second cycle as well.
(4 colours in Highres, 16 colours in Lowres)

3.5MHz/2*2bytes = 3.5MByte/s max transfer

In the A3000, A1200 and A4000, the CPU is clocked higher and the more modern 32bit variants could access the ram every cycle … if the Amiga would let them.
But while they implemented 32bit access to the ChipRAM, the CPU still can only access every second 280ns slot …

3.5MHz/2*4bytes = 7MByte/s max transfer

[No.3]

Originally Posted by Cyprian

- Bus bandwidth: 3.5MB/s for the CPU, 3.5MHz (1.7MHz for the CPU) 16bit memory bus

come on Cyprian, we had this topic last October... memory bus is running with 3.5 MHz and even if the CPU can access the memory only every 2nd cycle the memory bus is still running with 3.5 MHz.

[Cyprian]

@Gorf described it correctly.

Originally Posted by No.3

come on Cyprian, we had this topic last October... memory bus is running with 3.5 MHz and even if the CPU can access the memory only every 2nd cycle the memory bus is still running with 3.5 MHz.

you're right, I have to correct my self. Anyway the outcome is the same - bus bandwidth for the CPU.

Originally Posted by TEG

OK, I got it : 7MHz / 2 = 3.5 MHz (half timeslots for the chipset and half for the cpu in Lowres).

But now I have a problem with the 1.7 Mhz affirmation. The 68000 needed 4 cycles to fetch the memory. So 3.5 Mhz / 4 = 0.875 Mhz no ?

actually , this is 1.7 million memory slots available for the CPU.

The Amiga chipsets are clocked with 3.5MHz, Agnus needs two bus clocks for the CPU access (the first - for the address latches from CPU, the second is a phase where data is finally transferred), therefore this is 1.7 million memory slots available for the CPU (that 1.7MHz is just mental shortcut).

Moreover, If I'm not wrong, an access to the hardware registers is still 16bit, even in A1200/A4000

Originally Posted by TEG

Do you have a link for this 3.5MHz A4000 bus speed ?

some tests you can find below:
https://groups.google.com/g/comp.sys...m/aMw2s4s1dWQJ

A4000/040:
chip $001B8000 readw 860.9 ns normal 2.3 * 10^6 byte/s
chip $001B8000 readl 861.0 ns normal 4.6 * 10^6 byte/s

chip $001B8000 writew 863.0 ns normal 2.3 * 10^6 byte/s
chip $001B8000 writel 862.7 ns normal 4.6 * 10^6 byte/s


A4000/030:
chip $00028000 readw 538.9 ns normal 3.7 * 10^6 byte/s
chip $00028000 readl 573.3 ns normal 7.0 * 10^6 byte/s

chip $00028000 writew 575.0 ns normal 3.5 * 10^6 byte/s
chip $00028000 writel 574.8 ns normal 7.0 * 10^6 byte/s


https://eab.abime.net/showthread.php?p=1519982
Fast Ram:
40) A4000 64MB mobo fast ram = 11.5MB/s read, 6.1MB/s write


Is Zorro3 slot for graphics card? I see that it is not so fast also:
36a) A4000 ZorRam 256MB Zorro3 ram = 7.8MB/s readm, 4.4MB/s writem

Just for example, 640x512 8bit in 50Hz needs 16MB/s transfer

[sokolovic]

Originally Posted by Promilus

If you can prove that A1200 wasn't produced and sold after 94 then I'll admit I was wrong. If you can't prove that... who is the one lacking objectivity here?
Lifespan of hw is usually defined as time between it's debut up to the end of production and/or official support. Not my fault your definition is different.

I can prove you that Master System was produced and sold well after 2000.
The A1200 was Commodore flagship between 1992/1994. Escom ones were nothing but a way to grab some money with an old IP after everyone left the Amiga boat.

[Cyprian]

Originally Posted by Gorf

Unlike the ST we only got 7Mhz...

ah, true

Originally Posted by Gorf

To be fair: the A4000 came with FastRAM, which gives you fantastic 11.5MB/s read, 6.1MB/s write

Originally Posted by Gorf

AFAIK the Jaguar was not designed in-house ....
(Maybe Commodore should have hired some external developers to help them...)

yes and no, Atari hired Flare guys. They together created the Panther and its successor - the Jaguar. The question what was from Flare (pre-Atari), what was from Atari (pre-Flare), what was designed together.

My guess is that concept of the DSP / the GPU was from Flare (pre-Atari). The Object Processor was from Atari (pre-Flare) - actually it was based on Atari 7800 Maria's Display List which was done by - external company GCC .
The blitter looks like a much improved of Atari BLiTTER.
The bus - if I remember correctly the Panther had Atari ST like bus, the Jaguar had a different, faster one.

[Gorf]

Originally Posted by Cyprian

yes and no, Atari hired Flare guys. They together created the Panther and its successor - the Jaguar. The question what was from Flare (pre-Atari), what was from Atari (pre-Flare), what was designed together.

Somewhere I read this goes back to "Blossom" for the Atari-Transputer project....

The Object Processor was from Atari (pre-Flare) - actually it was based on Atari 7800 Maria's Display List which was done by - external company GCC .

Ah right .... GCC the company Jack did not want to pay after he bought Atari until they sued.

The concept of display lists itself is of course based on Jay Miners ideas and was also part of the lawsuit between Commodore and Atari, because one could argue the Copper-lists are a similar concept as in the 2600 chipset.
(But I would disagree)

[Cyprian]

Originally Posted by Gorf

Somewhere I read this goes back to "Blossom" for the Atari-Transputer project....

good point,
worthy of deeper analysis

Originally Posted by Gorf

Ah right .... GCC the company Jack did not want to pay after he bought Atari until they sued.

the story is not so simple,
Warner in early 1984 (before selling 51% of its assets to Tramiel) ordered a new console A7800 from GCC, but was reluctant pay them.
After Tramiel's acquisition of Atari in July 1984 (actually, Warner still co-owned Atari and had a 25% stake), Warner tried to force Tramiel to repay GCC. Because Tramiel was focused on his own product - RBP (Rock Bottom Price) - the ST, and really only had his money (a 'small' family budget) for that product, so he asked Warner for the loan. Negotiations were long (beginning in late 1984 to 1985) but were successful and Atari released 7800 in 1986

Originally Posted by Gorf

The concept of display lists itself is of course based on Jay Miners ideas and was also part of the lawsuit between Commodore and Atari, because one could argue the Copper-lists are a similar concept as in the 2600 chipset.
(But I would disagree)

In fact, all these ideas have noting in common other than the name
Atari 2600 had no Display List.
Atari 400/800 Display List - definition of graphic/text mode for incoming scanline, called once per line.
Atari 7800 DL/Jaguar OP - the list of sprites for incoming scanline, includes branches, called as many as there are sprites in the line.
Amiga Copper - the list of hardware register modifications, called many times per line.

[Gorf]

Thanks for the clarification!