Was anyone else disappointed with the A1200?

[Gorf]

Originally Posted by hammer

Amiga's F-29 Retaliator port used hardware acceleration. Amiga blitter has adder functions.

The difference between Amiga and ST in polygon drawing speed is more pronounced, when the CPU needs to work on something else - in this case game-logic and calculations.
In this case, the Blitter is a real advantage ... as long as it does not block memory access for the CPU.

When the transformations are pre-calculated, as in most demos, the advantage shrinks down, as you also need the CPU to set up the Blitter operations - so the difference between setting up the Blitter plus let it do the work and using the CPU directly to draw the polygons is quite small in many use cases.

Amiga's Dread is faster due to blitter accelerated C2P and displaying more than 16 colors due to hardware sprites and copper.

Not polygon based...
Dread is also available for the ST and it is not much slower there - almost the same frame rate.

[Thomas Richter]

Originally Posted by hammer

Amiga blitter has adder functions.

Nope. Unless you call an XOR a 1-bit addition.

[Gorf]

Originally Posted by hammer

An A2000 owner with full 32-bit 040 accelerator, the entire machine and CPU accelerator are dumped for A4000/040 side grade or A1200 side grade and buy another 040 accelerator. 040 is usually socketed. Does Commodore sell unpopulated A3640 card? Does Commodore sell A4000s with unpopulated CPU/FPU? Selling A2000 to another person doesn't expand AGA's user base. A4000's motherboard is not compatible with A2000's case.

Yes this was a sad move from Commodore:
An AGA motherboard upgrade in e.g. 93 for the A2000 and A3000(T) would have sold extremely well ... and it would not have hurt the sales of new machines much, since most owners of such machines would rather take the RTG-card option (as I did myself) than buying an AGA machine.

And it would have helped to push the AGA install base, which in turn would have been an incentive for developers to go this route.

[TEG]

Originally Posted by Gorf

The IRIS can move the flat shaded model in all angles in real-time.

Yeah, I saw an IRIS demonstration for real at the time, at a fair. Believe me, it had nothing in common with an Amiga. The demonstration I saw was using a big ball as a controller and when you turned it, the shape at the screen (a cone if I remember well) was recalculated and changed it shape and position in real time.

The reaction time was instantaneous, 100% smooth whatever the speed you moved the controller and the object covered almost all the screen. It was just another planet.

[hammer]

Originally Posted by Gorf

Yes this was a sad move from Commodore:
An AGA motherboard upgrade in e.g. 93 for the A2000 and A3000(T) would have sold extremely well ... and it would not have hurt the sales of new machines much, since most owners of such machines would rather take the RTG-card option than buying an AGA machine (as I did myself)

And it would have helped to push the AGA install base, which in turn would have been an incentive for developers to go this route.

AGA benefits Video Toaster 4000 and Scala MM400. RTG wasn't matured during 1992 to 1993 time period.

Commodore promised its RTG solution against GVP's EGS.
CGX was released in 1995 and there's no Commodore on creating FUD against it.

[hammer]

Originally Posted by Thomas Richter

Nope. Unless you call an XOR a 1-bit addition.

Related topic in https://eab.abime.net/showthread.php?t=77162 and https://github.com/Kalmalyzer/subpixel-blitter-line

[Gorf]

Originally Posted by hammer

AGA benefits Video Toaster 4000 and Scala MM400. RTG wasn't matured during 1992 to 1993 time period.

Commodore promised its RTG solution against GVP's EGS.
CGX was released in 1995 and there's no Commodore on creating FUD against it.

Not sure what you are trying to say ... seems missing my point.

3rd party gfx-cards for big box amigas where selling quite well in these years, even if the RTG software was mostly proprietary (EGS, Probench, PicassoRTG) and lacking features and compatibility.
Multiple vendors were fighting over a very small install base and were developing quite astonishing products ... and all of them were outlasting Commodore.

This shows that big-box-owners did hold on to their systems for quite some years - in the dire situation Commodore was in, offering a motherboard upgrade would generate some highly needed revenue - maybe even by just giving GVP or others a license to do so.

Selling my A3000 at a loss to get an A4000 was no option for me.
Also getting a slower and less expandable A1200 was no option.
And the A4000 was missing SCSI, which would have meant a new controller, to keep my SCSI equipment.
I assume most big-box-owners were in a similar situation...

[oscar_ates]

I think main culprit was Irvin Gould who messed things up for Amiga. He hired mehdi and for a while they even thought to drop amiga and become a pc company. Got slapped back by dell, gateway, clones.

[TEG]

Originally Posted by oscar_ates

I think main culprit was Irvin Gould who messed things up for Amiga. He hired mehdi and for a while they even thought to drop amiga and become a pc company. Got slapped back by dell, gateway, clones.

I can be wrong but in the mind of Gould, the Amiga was certainly a product, no more no less.

What is your source for the PC only company temptation?

[Bruce Abbott]

Originally Posted by Gorf

Such claims as in this ad lead to high expectations, which might then in turn lead to a disappointment after purchase - all without a need to do so, since the gfx powers of the Amiga were amazing ... just not really 3D.

I disagree. The whole point of the advert was to show that the A500 was not just a 'toy', ie. a games computer. When they said "dazzling 3D graphics manipulation and animation powers" they weren't talking about 3D games.

When Eric Graham created The Juggler animation in 1986 we were all blown away by it. Commodore thought he must have used a mainframe computer to render it, but he actually did the whole thing on Amiga 1000 using software he wrote by himself. It generated so much interest that in 1987 he released Sculpt 3D so anybody with an Amiga could do the same. This spurred the development of many other 3D rendering and animation programs, and also segued nicely with the Amiga's ability to produced broadcast standard video and (with a genlock) combine computer generated images with video from a VCR or camera. If you had an A500 then you could be a part of that.

Wikipedia classifies Agnus, Denise and Paula as the "Chipset" - And Gary, CIAs (and DMAC, Buster ...) as Custom Chips.
They all were dedicated chips for the Amiga, since you could not use the Amiga CIAs in an C64
https://en.wikipedia.org/wiki/Amiga_Original_Chip_Set
https://en.wikipedia.org/wiki/Amiga_custom_chips

Angus Denise and Paula were 'the Amiga chipset', but the A500's 'dedicated chipset' included Gary. The CIAs were not dedicated chips no matter what anybody (incorrectly) says to the contrary.

Retro Isle: Commodore 1571 troubleshooting guide

U20 CIA MOS 6526A (2 MHz) -or - MOS 8520 (2 MHz) - used only for fast serial access.

In the photo of the 1571 PCB in that link you can see an 8520 on the far right.

But this is all just technical nitpicking, the kind of stuff that only insufferable nerds like you and me pick up on (and can't agree on even among ourselves).

You have more of a point about the spelling mistakes, though they were not at all uncommon back in the days before sophisticated spell* checkers were the norm. We see plenty of them today too on high profile news sites etc., too due to writers relying too much on the spelling checker to alert them.

* an accepted usage of the word 'spell', despite the fact that as a noun it means "words held to have magic power", or "a state of enchantment" or "a strong compelling influence or attraction" - IOW all about magic and nothing to do with spelling.

[AnimaInCorpore]

Originally Posted by hammer

Amiga's F-29 Retaliator port used hardware acceleration. Amiga blitter has adder functions.

Besides screen clearing I see no other advantage of using the Blitter here.

Originally Posted by hammer

Amiga's Dread is faster due to blitter accelerated C2P and displaying more than 16 colors due to hardware sprites and copper.

The game performance on the Amiga mainly profits from Fast-RAM systems where the Blitter can run in parallel to the CPU. OTOH the Atari ST has an advantage with the bitplane format where you can do C2P quite effectively using the CPU. Also Copper like "colour enhancements" can be done on the Atari ST as well.

[hammer]

Originally Posted by Gilbert

Was anyone else disappointed with the A1200?

Most Amiga users and magazines seemed to be very happy with the A1200 when it came out. I wasn't at all, and a look at the first games pretty much ended my association with Amiga gaming. I just saw the same games with more colours and a bit smoother. There was no wow factor. After that I stuck with the Amiga 500 (with half meg memory expansion) and my Super Famicom (Jap SNES).

Here's what Commodore got wrong in my opinion

1. Too much focus on creating higher-res screen modes with more colours (and also making the blitter work in these different screen modes) and not enough on enhancing gaming(8 or maybe 16 sprites when the comparitively old Megadrive and SNES could manage 64 and 128 respectively). It's a bit like the original Amiga - yes it can display 4096 colours on screen, but the majority of the games for the system were 16 colours (Albeit some had added some Copper magic) and most didn't even run at 50/60 fps. That was fine back in 1985 but 7 years(!) later you expect a significant upgrade.

2. There was a mild improvement to dual playfield mode. Great!... when the SNES had 5(?) playfields and could scale and rotate whole screens. Commodore seemed to have no sense they were competing here....

2. Sound chip needed 6 channels to get a decent track playing with sound effects. Again SNES and Megadrive have 6 channels each. Using the same sound chip from 1985 was ridiculous!

3. Like the original Amiga, if you wanted to get a good number of objects on screen with a lot of colours and scrolling, you had to spend ages using hardware tricks or specific techniques. Time = money and developers aren't going to want to spend 2 years making an arcade quality game on the A1200 when simpler systems exist....

I do have a CD32 now, but it's not very impressive from a technical point of view, even the mighty Banshee is bettered on both the SNES and Megadrive. The reason I like it is because it offers something a bit different and it's an Amiga It's fairly obvious it had no hope of competing long term. I just find it hard to see what Commodore was thinking with the AGA architecture??

Quake 1's timedemo 1 benchmark

[ Show youtube player ]
PiStorm32 Lite Emu68 (June 2023 build) with RPi 4B's ARM Cortex A72 @ 1.8Ghz (three ARM instructions per cycle) with Commodore's AGA display
Result: 47.23 fps for 320x200

It's about 49 fps in recent Emu68 buids.

From https://old.vgamuseum.info/benchmark...rendering.html
With Athlon XP 2200+ (1.8 Ghz, three X86 instructions per cycle), Soltek SL75-KAV (Via KT133A), 256MB SDR CL3

When both machines has similar class 1.8 Ghz CPUs, AGA will beat all ISA based SVGA cards at 320x240 resolution level.
Best ISA SVGA card reached 36.2 fps.


A high CPU performance removes the CPU from the bottleneck factor, hence it's pure display and CPU to video bus I/O capability.

For 50 hz PAL, 25 fps is needed.
For 60 hz NTSC, 30 fps is needed.

[hammer]

Originally Posted by AnimaInCorpore

Besides screen clearing I see no other advantage of using the Blitter here.

The game performance on the Amiga mainly profits from Fast-RAM systems where the Blitter can run in parallel to the CPU.

Nope, 68000's IPC is low i.e. an instruction takes many clock cycles to compete while Amiga's custom chips have the opportunity execute its instructions. Blitter includes a barrel shifter feature. Wake me up when 68000 @ 8 Mhz delivers 3.5 MIPS.

68020 has a higher IPC (and includes hardware barrel shifter) when compared to 68000, hence adding Fast RAM for A1200 has greater benefits.

Amiga's custom chips are designed to patch 68000's low IPC and incompetent multimedia capability.

Amiga Dread's line draw is Blitter accelerated, hence conserving CPU resource.

Originally Posted by AnimaInCorpore

OTOH the Atari ST has an advantage with the bitplane format where you can do C2P quite effectively using the CPU. Also Copper like "colour enhancements" can be done on the Atari ST as well.

Atari ST's race the beam palette change is CPU driven via interrupt timer B. Amiga's race the beam palette change function is not govern by the CPU i.e. Copper's task, hence conserving CPU resource.

CPU upgrade affects Atari ST's game backwards compatibility more than Amiga since the Copper isolates most of the graphics effects from the CPU.

Amiga's 8 hardware sprites has separate color palette and the Copper can be applied on hardware sprites. Dread's main player gun graphics are hardware sprites on the Amiga. Copper can be use to move hardware sprites. Copper usage conserves CPU resource.

Atari ST doesn't have hardware sprites.

If only the Copper was updated into vertex co-processor. There's 3DO.

[hammer]

Originally Posted by Bruce Abbott

I disagree. The whole point of the advert was to show that the A500 was not just a 'toy', ie. a games computer. When they said "dazzling 3D graphics manipulation and animation powers" they weren't talking about 3D games.

When Eric Graham created The Juggler animation in 1986 we were all blown away by it. Commodore thought he must have used a mainframe computer to render it, but he actually did the whole thing on Amiga 1000 using software he wrote by himself. It generated so much interest that in 1987 he released Sculpt 3D so anybody with an Amiga could do the same. This spurred the development of many other 3D rendering and animation programs, and also segued nicely with the Amiga's ability to produced broadcast standard video and (with a genlock) combine computer generated images with video from a VCR or camera. If you had an A500 then you could be a part of that.

The whole point with selecting weak IPC 68000 is 32-bit programming model for 32-bit desktop OS road map with 68020. Think of the half baked 386SX @ 7.1 Mhz when 386 didn't exist in 1983.

A pure game console has little consideration for 32-bit desktop OS road map, hence SNES direction.

When EGA is not enough, Amiga's 4096 color palette targets similar era workstation class' 4096 color palette solution e.g. IBM PGC (1984) and NEC uPD7220 for PC-98 (1982). Amiga 1000 was "workstation class for the masses" attempt. Commodore made it cheap for the masses with the A500. "Workstation class for the masses" ideology was lost by Commodore's post-A500 road map.

Today's GpGPU and X86 companies has "workstation class for the masses" ideology via assimulated or stolen or influenced by workstation IP.

DEC couldn't sustain a legal fight against Intel, hence assimulated DEC's CPU teams. Samsung couldn't continue Alpha R&D since they are missing CPU designers.
SGI couldn't sustain a legal fight against NVIDIA hence assimulated SGI's graphics team.
Microsoft hired the VMS designer.
AMD assimulated SGI rebels i.e. ArtX.
NVIDIA assimulated SGI rebels i.e. 3DFX.
AMD assimulated DEC rebels i.e. NexGen.

X86 collective assimulated the best of workstation CPU and graphics personnel and their skillset is gained.

[Gorf]

Originally Posted by hammer

If only the Copper was updated into vertex co-processor. There's 3DO.

No use for that without VRAM (bandwidth) and chunky pixel format.

Even some simple 3D trick like Mode 7 on the SNES requires a chunky mode
(and even the SNES is actually wasting halve of the bandwidth in this mode with 8bit pixels and 16bit fetches)

[Gorf]

Originally Posted by Bruce Abbott

When they said "dazzling 3D graphics manipulation and animation powers" they weren't talking about 3D games.

Neither was I.

[AestheticDebris]

Originally Posted by hammer

A pure game console has little consideration for 32-bit desktop OS road map, hence SNES direction.

That doesn't really make any sense. The SNES was.going to have a 68000 but Nintendo were considering NES compatibility out of the box for some time and so certain aspects of it's design, including CPU choice, were instead chosen to make that more feasible.

If 68k wasn't suitable for consoles, why would Sega use it?

Originally Posted by Gorf

Even some simple 3D trick like Mode 7 on the SNES requires a chunky mode
(and even the SNES is actually wasting halve of the bandwidth in this mode with 8bit pixels and 16bit fetches)

The SNES doesn't have chunky pixels, although you can sort of fake it in Mode 7 by filling each tile with a single colour and then zooming the display so each uses 1*1 (or 2*2.or whatever) pixels. Then changing any given "tile" in the map effectively plots a "pixel" and it's a bit like having a chunky bitmap format. The tiles themselves are stored in a planar like format, however.

[Bruce Abbott]

Originally Posted by Gorf

Neither was I.

Then I'm confused. Was the Amiga not capable of 'dazzling 3D graphics manipulation and animation'? Why shouldn't they have promoted it, because a machine costing over 20 times more could maybe do it better?

The extent Amiga fans will go to to denigrate their girlfriend is amazing. Never a good word to say about her.

[ Show youtube player ]

[Thorham]

Originally Posted by Bruce Abbott

The extent Amiga fans will go to to denigrate their girlfriend is amazing.

Amiga doesn't mean girlfriend

[Bruce Abbott]

Originally Posted by hammer

The whole point with selecting weak IPC 68000 is 32-bit programming model for 32-bit desktop OS road map with 68020. Think of the half baked 386SX @ 7.1 Mhz when 386 didn't exist in 1983.

The 68000 wasn't 'half baked' when it was introduced in 1979. In fact it was considered overambitious, which is one reason IBM chose the 8088 instead. For sure Motorola intended it to eventually go fully 32-bit, but it would be many years before they managed to achieve it. Meanwhile Apple used the 68000 in the Mac without any thought of a 32-bit future.

386SX wasn't 'half baked' either. It was a deliberate (and successful) attempt to bring 32-bit technology to the masses. The idea behind the 386SX was that it could be fitted to cheaper 286 motherboards without a major redesign of the board or chipset.

A pure game console has little consideration for 32-bit desktop OS road map, hence SNES direction.

Duh.

When EGA is not enough, Amiga's 4096 color palette targets similar era workstation class' 4096 color palette solution e.g. IBM PGC (1984) and NEC uPD7220 for PC-98 (1982).

There was a lot more to the Amiga's graphics than just a 4096 color palette.

The Apple IIGS had 4096 colors in 1986. By 1990 the Amstrad CPC and Atari ST had it to (and the C65 would have had it). But none of them had HAM mode and a copper that could put all 4096 colors on screen at once.

IBM PCG was ridiculously expensive and had very poor bitmap animation capability - not even worth mentioning in the same space as the Amiga.

PC-98 was NEC's proprietary MS-DOS machine that was only sold in Japan. It too was very expensive and a nonentity in the West.

Amiga 1000 was "workstation class for the masses" attempt. Commodore made it cheap for the masses with the A500.

True.

"Workstation class for the masses" ideology was lost by Commodore's post-A500 road map.

False. '"workstation class for the masses" attempt' is not the same as actual "workstation class". The A1000 was only a 'toy' workstation, ie. a home computer that people could pretend was a workstation. Real workstations were designed for maximum productivity to get commercial jobs done.

Commodore's actual attempts at workstation class machines included the A2500 with 68020 and then 68030, and the A3000, all of which were designed with UNIX in mind. But the engineers didn't have the skills to pull it off, and were always one step behind SGI. That 'road map' had to die before it killed Commodore.

Today's GpGPU and X86 companies has "workstation class for the masses" ideology via assimulated or stolen or influenced by workstation IP.

Today, 35 years later? How is this relevant? It was going to happen anyway whether they 'stole' workstation IP or not.

SGI couldn't sustain a legal fight against NVIDIA hence assimulated

SGI made overpriced proprietary machines that couldn't compete against the PC. They were toast even without any legal fights. Once people discovered that they could do the same stuff on their PCs (even if a bit slower and clunkier) it was all over for SGI.

Here's the truth. In 1981 the desktop computer industry was desperate for a way to be taken more seriously. That came in the form of IBM with their PC. The industry seized on the PC as their holy grail and the rest was history. Anything else would either be stomped on or assimilated, because they had chosen the 'one' architecture that would be their roadmap for the future, wherever that might lead. Nobody could buck it for long - not even IBM itself. Commodore valiantly fought against that for almost a decade, constantly putting faith in a different architecture that was bound to fail eventually - simply because it wasn't IBM compatible.