Was anyone else disappointed with the A1200?

[TCD]

Originally Posted by Cris1997XX

Damn, seeing you all bite into hammer's bait is really pitiful...

I think that at least Bruce might actually be into that kind of stuff

[emiespo]

Originally Posted by Bruce Abbott

But hey, perhaps you are right and it wouldn't be any more complex. Unfortunately - unlike today's armchair engineers - Jay Miner didn't think to do all it in packed pixel mode. Strangely nobody else did either. I guess they just weren't as smart as Amiga fans of today - probably due to all that tetraethyl lead they were breathing in back then.

Unlike angry armchair engineers, this is what wrote multiple times: discuss the reasons why it was done the way it was done, I mentioned cost saving, and simplifying the circuitry and debugging time is cost saving.

By the time AGA came around I don't remember many machines with a silly 8 bitplanes mode… hardware development tools were also slightly better. Also packed 4 bit modes were fairly common.

I can think of more copper optimizations too, but this doesn't mean what we got was inadequate. It was far better than anything else out there. And it actually worked, unlike so many bright ideas that turned out to be impossible to implement. The AGA chipset was complex enough already, without loading it up with feature creep that would delay its introduction even more.

Of course, no sane armchair engineer would come up with the same copper, blitter and Paula from 1985 seven sound years later. I presume in this perspective even the C64GS was a great idea… I really don’t get why Commodore is not around anymore with such “far better” products.

Simple question: have you ever used the Workbench in an 8 bit (planes… ) VGA-only mode? With the blitter and cpu gasping for some breath in accessing the overwhelmed chip-ram?

Even with this in mind, the 1200 was relatively cheap for what it offered, despite losing in sheer speed with any PC clone similarly priced (add HD, accelerator and a Commodore 1960* to the 1200), the A4000 was ridiculously overpriced for what it offered.


* AGA had weird display modes, they’d look all good only on that custom multisync monitor from C=

[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??

https://archive.computerhistory.org/...-05-01-acc.pdf
1993 published wholesale 1993 to 1997 price guide for chip components. An interesting read.

For Q1 1993,
68EC020-16 PQFP has $15 price.

68EC020-25 PQFP has $18 price. A very minor price increase for a smoother Wing Commander CD32 experience. David Pleasance's and major 3rd party developer's arguments are correct. Mehdi Ali's argument position is shit.

68EC030-25 PQFP has $34 price.

68030-25 CQFP has $54.50 price. $20 extra just for MMU.

A1200 and CD32 have 68EC020-16-rated CPUs with minor downclock to 14 Mhz. 68EC020-25 could survive a minor 28 Mhz overclock when 3rd party Amiga CPU accelerators are doing it.

Commodore's product segment tactics are to funnel any die-hard Amiga fans who wanted faster than A1200's 68EC020-16 into higher profit margins A4000/EC030 which cost like 486SX-33 PC clones e.g. Gateway 2000's.

The Amiga's product stack is missing Apple's Quadra 605 (68LC040)'s $975 price point.

Meanwhile...

68EC040-25 has $91.00 price. 68LC040's price would be close to this price.

AM386DX-40 PQFP has $48.25 price.

386DX-25 PQFP has $60.90 price.

80486SX-25 PQFP has $89.00 price.

PS; 68EC040 and 68EC060 will not be compatible with the Amiga due to missing MMU. For the 68040 bus, it needs the MMU to mark the Chip RAM address range as non-cacheable. 68030 has the signal pins for non-cacheable when accessing the Chip RAM address range.

Mass-produced A1200/CD32 with 68EC020 @ 25 Mhz and 2MB 32-bit Fast RAM would be "power without the price" against PC's i386DX-25 competition, but AMD's 386DX-40 is a strong competitor. A jumper for 28Mhz minor overclock is on the end user.


----------------
$799 AUD A1200 with 68EC020 @ 25Mhz CPU... I'll buy it in 1993 and ditch my A3000(030 @ 25Mhz).

[hammer]

Originally Posted by TCD

I think that at least Bruce might actually be into that kind of stuff

Don't defend Mehdi Ali's argument. Mehdi Ali's argument position is shit.

[hammer]

Originally Posted by paul1981

Some would say like artificial intelligence.

Bruce Abbott is defending Mehdi Ali's argument.


https://archive.computerhistory.org/...-05-01-acc.pdf
1993 wholesale price guide for chip components. An interesting read.

For Q1 1993,
68EC020-16 PQFP has $15 price.

68EC020-25 PQFP has $18 price. A tiny price increase for smooth Wing Commander CD32.


Mehdi Ali's argument is shit along with anyone who defends Mehdi Ali.

[hammer]

Originally Posted by emiespo

Unlike angry armchair engineers, this is what wrote multiple times: discuss the reasons why it was done the way it was done, I mentioned cost saving, and simplifying the circuitry and debugging time is cost saving.

Commodore management caused time wasting, resource misdirection, large-scale corruption, and the loss of proven good engineers.

Originally Posted by emiespo

By the time AGA came around I don't remember many machines with a silly 8 bitplanes mode… hardware development tools were also slightly better. Also packed 4 bit modes were fairly common.

From [ Show youtube player ]
1. ET4000W32i ISA card handles 800x600 with 16-bit color Windows GDI just fine.

2. ET4000W32i ISA card handles StarCraft's 640x480p 256 colors just fine when the Windows GDI display setting is 640x480p 16-bit hi-color. StarCraft used DirectDraw API.

Gaming PC won for valid technical and cost vs performance reasons.

Commodore's core revenue demographics are in games.

Originally Posted by emiespo

Of course, no sane armchair engineer would come up with the same copper, blitter and Paula from 1985 seven sound years later. I presume in this perspective even the C64GS was a great idea… I really don’t get why Commodore is not around anymore with such “far better” products.

3DO says Hi... 3DO was 36 channel DMA beast.

https://en.wikipedia.org/wiki/Amiga_Ranger_Chipset
Commodore's original Los Gatos Amiga team shutdown comes to bite back. I rather keep the original Los Gatos Amiga team and fire PC advocates in Commodore!

Originally Posted by emiespo

Simple question: have you ever used the Workbench in an 8 bit (planes… ) VGA-only mode? With the blitter and cpu gasping for some breath in accessing the overwhelmed chip-ram?

I tried to use my A1200 with TF1260 like my 386DX-33/ET4000-based PC including AGA's Double NTSC 640x400p 256 color Workbench.

AGA's Double NTSC 640x400p 64 color Workbench is okay.

Originally Posted by emiespo

Even with this in mind, the 1200 was relatively cheap for what it offered, despite losing in sheer speed with any PC clone similarly priced (add HD, accelerator and a Commodore 1960* to the 1200), the A4000 was ridiculously overpriced for what it offered.

* AGA had weird display modes, they’d look all good only on that custom multisync monitor from C=

Alice's Blitter is still 16-bit with OCS/ECS timings.

AGA is reasonable for a fast +50 fps 320x200/256 with 256 color frame buffer for 2.5D/3D gaming, but it's missing a fast object manipulation processor.

The price difference between 68EC020 25 Mhz and 16 Mhz is very minor.

[Bruce Abbott]

Originally Posted by emiespo

By the time AGA came around I don't remember many machines with a silly 8 bitplanes mode… hardware development tools were also slightly better. Also packed 4 bit modes were fairly common.

It wasn't silly. Jay Miner designed the original chipset with 6 bitplanes in 1983. By 1987 he had extended it to 7 bitplanes (Ranger). Only one more bitplane to go! The OS always supported 8 bitplanes, so 8 bitplanes in AGA was natural.

Of course, no sane armchair engineer would come up with the same copper, blitter and Paula from 1985 seven sound years later.

...unless they were designing something intended to be compatible with the Amiga. No 'sane' engineer would have come up with CGA, EGA or Hercules graphics either, but they were putting that 'silly' stuff into PC graphics cards. Why? For compatibility.

I presume in this perspective even the C64GS was a great idea… I really don’t get why Commodore is not around anymore with such “far better” products.

C64GS was not a great idea.

Simple question: have you ever used the Workbench in an 8 bit (planes… ) VGA-only mode? With the blitter and cpu gasping for some breath in accessing the overwhelmed chip-ram?

Why would I? 640x256 in 8 colors is plenty enough for me, and my TV can't handle VGA scan rates on the composite input.

ChipRAM is no more 'overwhelmed' in 256 colors in VGA-only mode than OCS is 16 colors. But of course you want to max everything out and then complain about it being too slow. So just having 256 colors isn't enough, it also has to be 640x480 resolution, right?

This is where Commodore made their mistake. They should have done what Atari did and limited the number of colors at high resolutions. Want VGA at 640x480? No more than 16 colors for you! Then fans couldn't complain about it being slow, because they didn't have it!

Even with this in mind, the 1200 was relatively cheap for what it offered, despite losing in sheer speed with any PC clone similarly priced (add HD, accelerator and a Commodore 1960* to the 1200)

Some people just don't get it. If you want equivalent hardware you have to pay equivalent money - more when the hardware is less commoditized.

The advantage of the A1200 was that for that much lower price you got a usable computer system, with access to the Amiga ecosystem. A PC was useless without a hard drive and monitor etc., so you had to buy all that stuff with it. The A1200 could be used on your existing TV, and ran software directly off floppies so you didn't need a hard drive, and games were designed to work on a stock machine so you didn't need a faster CPU and more RAM.

That's how it was done with home computers from the beginning. The cheaper systems got there by having a design didn't use as much hardware. Of course that meant they had limitations - like the ZX Spectrum which used your own cassette recorder, only had a single bitplane display with character attribute color, and no joystick port. Pared down to the bare minimal to do the job. Or the TI-99/4a, which was a 16 bit computer with only 256 bytes of 16 bit RAM. The only other RAM was on the the other side of the video display processor, where it could only be accessed as data via an I/O port.

If we want to talk about 'silly' designs, there are plenty of examples that make the Amiga's designers look like geniuses (which they were).

the A4000 was ridiculously overpriced for what it offered.

The A4000 was intended for professionals who didn't care about price. It wasn't overpriced if you wanted to put a Video Toaster into it because the equivalent A2000 system would cost about the same. And it was still cheap compared to the professional video equipment you had it hooked up to.

* AGA had weird display modes, they’d look all good only on that custom multisync monitor from C=

Another thing Commodore shouldn't have offered. If people wanted to create their own weird video modes then fine. And guess what, they did! Even enjoyed doing it.

[hammer]

Originally Posted by Bruce Abbott

It wasn't silly. Jay Miner designed the original chipset with 6 bitplanes in 1983. By 1987 he had extended it to 7 bitplanes (Ranger). Only one more bitplane to go! The OS always supported 8 bitplanes, so 8 bitplanes in AGA was natural.

I agree that 8-bit plane support is close.

1983 = 5 bit planes, some original NTSC A1000 is missing EHB mode. HAM has 6-bit planes.

1985 = 6 bit planes. PAL and newer NTSC A1000s have 64-color EHB mode.

1987 = 7 bit planes.

For almost every two-year interval, Jay Miner's team is increasing the bit planes. 1989 could have 8-bit planes and this would be ready for SNES's 1990 release and any PC fast VGA's 256 color modes e.g. ET4000AX's 1989 release.

I would be very happy with 1989-era A500 Rev 6A and 1990 A3000 with AGA. AGA platform would have enough time to build up a large install base against SNES's 1992 release in Commodore's core revenue European market.

After 8 bit planes, 1-byte (8 bit) and 2-byte (16-bit) packed pixels would be next e.g. 1991 and 1993 releases. 3DO was released in Q4 1993.

Originally Posted by Bruce Abbott

...unless they were designing something intended to be compatible with the Amiga. No 'sane' engineer would have come up with CGA, EGA or Hercules graphics either, but they were putting that 'silly' stuff into PC graphics cards. Why? For compatibility.

C64GS was not a great idea.

C65's 256-color R&D commitment is okay, but C64's target market wasn't a good idea.

Originally Posted by Bruce Abbott

Why would I? 640x256 in 8 colors is plenty enough for me, and my TV can't handle VGA scan rates on the composite input.

My C= 1084S couldn't handle VGA, but the C= 1942 monitor can handle it.

I would have purchased C= 1942 monitor since it has a similar use case as my Dell LCD monitor's 15 kHz and VGA/SVGA support.

For my A3000, I have PC SVGA (via Amber Flicker Fixer) and 1084S monitors. A3000's Amber Flicker Fixer enables some cheap PC VGA/SVGA monitors with superior dot pitch. PC VGA has trouble with PAL 50 hz, but it's okay with NTSC's 60 hz.

A basic frequency doubler from 15.6kHz to 31.1kHz would be nice. I don't care about Amber Flicker Fixer's de-interlacing with VRAM buffer when AGA has Double NTSC.

Originally Posted by Bruce Abbott

ChipRAM is no more 'overwhelmed' in 256 colors in VGA-only mode than OCS is 16 colors. But of course you want to max everything out and then complain about it being too slow. So just having 256 colors isn't enough, it also has to be 640x480 resolution, right?

This is where Commodore made their mistake. They should have done what Atari did and limited the number of colors at high resolutions. Want VGA at 640x480? No more than 16 colors for you! Then fans couldn't complain about it being slow, because they didn't have it!

AGA Lisa has memory access up to 28MB/s memory bandwidth from theoretical 56 MB/s but the raster is nearly useless without a matching object manipulator.

AGA Lisa has no problems displaying pre-baked frames. At least a dual 16-bit Blitter config.

[Bruce Abbott]

Originally Posted by hammer

3DO says Hi... 3DO was 36 channel DMA beast.

3DO says "Hey, look how overpriced I am!"

https://en.wikipedia.org/wiki/Amiga_Ranger_Chipset
Commodore's original Los Gatos Amiga team shutdown comes to bite back. I rather keep the original Los Gatos Amiga team and fire PC advocates in Commodore!

Thankfully you didn't get a say in the matter.

Perhaps you didn't know that the Amiga was going to have IBM compatibility even before Commodore bought it. Jay Miner even wanted it to look like a PC. It was only after they designed the A1000 and wanted to keep that form factor that he changed his mind.

AGA's Double NTSC 640x400p 64 color Workbench is okay.

It's more than OK. Most PC users at the time ran Windows in 16 colors because that's what it was designed for and the next up (256 colors) was too taxing.

Alice's Blitter is still 16-bit with OCS/ECS timings.

But 60% faster in AGA with 16 colors due to less DMA contention.

AGA is reasonable for a fast +50 fps 320x200/256 with 256 color frame buffer for 2.5D/3D gaming, but it's missing a fast object manipulation processor.

Whereas the PC was missing any 'object manipulation processor' and had to do it all with the CPU. With intelligent use of sprites and blitter the CPU load can be lightened significantly. Dread is a good example. Weapons are impressively detailed but use up almost no CPU time, and the HUD is rendered efficiently in 2D. That's how even the humble A500 gets a reasonable frame rate and still looks good. On a stock A1200 it really flies!

The trick to getting good performance out of the A1200 is not to ask too much from it. Do you really need 256 colors in hi-res, or are you just using them because you can?

OCS games were often limited to 16 colors for speed, perhaps with a copper gradient background that didn't look great due to the limited color palette. But AGA can do 16 colors in dual playfield, with an extra 2 or 4 colors in the background, plus super smooth copper gradients and large 16 color sprites for even more effects, all without running out of bandwidth or needing more blit time.

If you just do everything in 256 colors then you lose that efficiency. On a PC you have to do that anyway as working with fewer colors isn't an option (unless you go down to stinky 16 color EGA), but AGA isn't so limited.

[hammer]

Originally Posted by Bruce Abbott

3DO says "Hey, look how overpriced I am!"

It wouldn't be overpriced when the $699 price target is a desktop computer. Commodore's cost reduction specialty is needed.

The original A1000 and A500 needed the original Amiga team and Commodore's cost reduction specialty. Both halves made the original Amiga possible. Skilled personnel are the most important asset in a tech company.

Originally Posted by Bruce Abbott

Thankfully you didn't get a say in the matter.

Commodore is dead while Apple has a superior focus on their in-house Mac platform.

Originally Posted by Bruce Abbott

Perhaps you didn't know that the Amiga was going to have IBM compatibility even before Commodore bought it. Jay Miner even wanted it to look like a PC. It was only after they designed the A1000 and wanted to keep that form factor that he changed his mind.

1. 68000's big-endian is the opposite of PC X86's little-endian. The Transformer PC emulator is slow.

2. The Amiga has its game console origin when Amiga custom chips are not designed to be PC's modular design. When AGA was released, the existing full 32-bit 68020/030/040 CPU equipped with Amigas wasn't able to join the 256-color AGA target.

The same era full 32-bit 386DX-based PC clones can join 256-color VGA target with SVGA card upgrades.

Amiga's game console origin revealed itself when the Amiga chipset section wasn't PC-style modular.

Originally Posted by Bruce Abbott

It's more than OK. Most PC users at the time ran Windows in 16 colors because that's what it was designed for and the next up (256 colors) was too taxing.

Again, that's a flawed argument when the PC market is very large.

https://dosdays.co.uk/topics/Manufac...tseng_labs.php
By 1991, according to IDC, Tseng Labs held a 25% market share in the total VGA market.

From https://www.intel.fr/content/dam/doc...ual-report.pdf
Intel reported the following
1. In 1994's fourth quarter, Pentium unit sales accounted for 23 percent of Intel's desktop processor volume.
2. Millions of Pentiums were shipped.
3. During Q4 1993 and 1994, a typical PC purchase was a computer featuring the Intel 486 chip.
4. Net 1994 revenue reached $11.5 billion.
5. Net 1993 revenue reached $8.7 billion.
6. Growing demand and production for Intel 486 resulted in a sharp decline in sales for Intel 386 from 1992 to 1993.
7. Sales of the Intel 486 family comprised the majority of Intel's revenue during 1992, 1993, and 1994.
8. Intel reached its 6 to 7 million Pentiums shipped goal during 1994. This is only 23 percent unit volume.

By the end of 1994, Intel's Pentium PC install base crushed the entire Amiga install base of 4 to 5 million units!

Originally Posted by Bruce Abbott

But 60% faster in AGA with 16 colors due to less DMA contention.

640x400 is 4X from 320x200 pixels. AGA Lisa has up to 4X memory bandwidth improvements, but Alice's Blitter wasn't 4X scaled. Run out of time due to Commodore management's time wasting.

Again, Commodore didn't have a strong 256-color use case baseline until they decided to branch from AAA into rash job AGA.

Originally Posted by Bruce Abbott

Whereas the PC was missing any 'object manipulation processor' and had to do it all with the CPU.

My statement on "object manipulation processor" covers the CPU or Blitter. At the end of the day, it's about delivering a certain gaming experience and for what price.

The PC clone market is optimized for strong competition among PC vendors.

Originally Posted by Bruce Abbott

With intelligent use of sprites and blitter the CPU load can be lightened significantly. Dread is a good example. Weapons are impressively detailed but use up almost no CPU time, and the HUD is rendered efficiently in 2D. That's how even the humble A500 gets a reasonable frame rate and still looks good. On a stock A1200 it really flies!

Commodore didn't have a strong 1st party game studio for establishing relevant use cases and creating game-optimized SDK.

Dread/Grind's line skip tricks were implemented on the Sega Mega Drive.

Dread/Grind is a good "what if" when the Amiga platform has excellent game developers like on the Sega Mega Drive.

Originally Posted by Bruce Abbott

The trick to getting good performance out of the A1200 is not to ask too much from it. Do you really need 256 colors in hi-res, or are you just using them because you can?

640x400 is 4X from 320x200 pixels. AGA Lisa has up to 4X memory bandwidth improvements, but Alice's Blitter wasn't 4X scaled. Run out of time due to Commodore management's time wasting.

Originally Posted by Bruce Abbott

OCS games were often limited to 16 colors for speed, perhaps with a copper gradient background that didn't look great due to the limited color palette.

From the art asset rip, Elf Mania has 32 colors without Atari ST's color palette consideration.

Elf Mania needs player-control design changes. Elf Mania is a good tech demo.

Originally Posted by Bruce Abbott

But AGA can do 16 colors in dual playfield, with an extra 2 or 4 colors in the background, plus super smooth copper gradients and large 16 color sprites for even more effects, all without running out of bandwidth or needing more blit time.

That's not going beyond the Sega Mega Drive's 2D gaming experience.

A better AGA example is [ Show youtube player ]
Street Fighter II AGA tech demo. It uses Copper color tiling/zoning and 64-bit wide sprites for parallax background. Not many mainstream game programmers will use this optimization route i.e. it's hard.

A dual 16-bit Blitter is preferred. Alice's 16-bit Blitter is wasting half of the clock cycle opportunity.

For example, SNES's Mortal Kombat has more than 100 colors with parallax layers.

The low-cost market segment SNES has a strong 2D gaming experience.

Originally Posted by Bruce Abbott

If you just do everything in 256 colors then you lose that efficiency. On a PC you have to do that anyway as working with fewer colors isn't an option (unless you go down to stinky 16 color EGA), but AGA isn't so limited.

PC VGA has discrete video memory. There's VGA 320x200 16 colors mode i.e. Mode D. PC VGA is not limited by 320x200 256 colors e.g. PC Pinball Fantasies has a custom resolution mode.
The PC has VGA hardware features that are friendly for Doom-type games.

PC's "new 32-bit 2.5D/3D gaming experience" is beyond the low-cost segment SNES's strong 2D gaming experience.

You either have strong game developers or strong hardware.

Anyway, Doom's artists (Adrian Carmack and Kevin Cloud) used Deluxe Paint II for DOS to produce the artwork for the game. LOL

[hammer]

Originally Posted by Bruce Abbott

Some people just don't get it. If you want equivalent hardware you have to pay equivalent money - more when the hardware is less commoditized.

The advantage of the A1200 was that for that much lower price you got a usable computer system, with access to the Amiga ecosystem. A PC was useless without a hard drive and monitor etc., so you had to buy all that stuff with it. The A1200 could be used on your existing TV, and ran software directly off floppies so you didn't need a hard drive, and games were designed to work on a stock machine so you didn't need a faster CPU and more RAM.

Wholesale price for Q1 1993:

68EC020 25 Mhz = $18
68EC020 16 Mhz = $15

That's a minor price difference for a smoother Wing Commander CD32.

A1000+ AGA with 68EC020 @ 25 Mhz was canceled for ECS A1000Jr!

[sandruzzo]

AGA with 68EC020 @ 25 Mhz required faster and better chip mem, maybe that would have raised the price

[Cyprian]

Originally Posted by hammer

AGA Lisa has memory access up to 28MB/s memory bandwidth from theoretical 56 MB/s

how did you calculate that 56MB/s?

[sandruzzo]

Maybe 14mhz ram x 64 bit wide bus?

[Cyprian]

Originally Posted by sandruzzo

Maybe 14mhz ram x 64 bit wide bus?

ok, but the bus (and chipsets) is 3.5MHz 32bit

[pandy71]

Originally Posted by hammer

For Q1 1993,
68EC020-16 PQFP has $15 price.

68EC020-25 PQFP has $18 price. A tiny price increase for smooth Wing Commander CD32.

Mehdi Ali's argument is shit along with anyone who defends Mehdi Ali.

No one defend Mehdi Ali but you obviously never worked in large company producing mass selling goods... people trying to cut design by using 3..4 cents cheaper part and you suddenly propose 3$ - in large company those 3$ is multiplied immediately by at least 1 million and convince management to accept 3 million $ loss - good luck with this...

Originally Posted by hammer

Commodore management caused time wasting, resource misdirection, large-scale corruption, and the loss of proven good engineers.

Are you share owner? stock investor with loss on Commodore papers? then sue them...
Most of those you are blaming are probably from very long time decomposed... you could try if you wish to repeat something like this: https://en.wikipedia.org/wiki/Cadave...useskin=vector good luck with this...

Originally Posted by hammer

1. ET4000W32i ISA card handles 800x600 with 16-bit color Windows GDI just fine.

2. ET4000W32i ISA card handles StarCraft's 640x480p 256 colors just fine when the Windows GDI display setting is 640x480p 16-bit hi-color. StarCraft used DirectDraw API.

If you say so - all above chips are for VLB or for PCI bus and usually looks less shiny than competition namely S3 solutions or ATI MACH32.

I owned W32i with 486DX66 VLB (real DX50 overclocked to 66MHz with IDE on VLB - it was VERY TRICKY to make it work at all) and quickly this fast config become obsolete...

PCI killed Tseng...

Originally Posted by hammer

Wholesale price for Q1 1993:

68EC020 25 Mhz = $18
68EC020 16 Mhz = $15

That's a minor price difference for a smoother Wing Commander CD32.

A1000+ AGA with 68EC020 @ 25 Mhz was canceled for ECS A1000Jr!

Once again - big corpo works and thinks differently and for sure they usually don't care about customer needs - this is like Apple - "We know better what do you need and how to make you happy and if you don't agree with this then sod off".

3$ is large difference in mass production when you need to take risk and made pre-orders and this is more than just 3$ as you need to use faster memories or add cache or both... You need to take from pocket some money with risk of no profit... This how life works... In single peaces 3$ are fine, low risk 30..300$ loss well non one like to lose but hey - worst things happen everyday...

There is many ways to speedup Agnus/Allice but they need complex design and may ruin legacy compatibility - i can understand lack of will to take risk...
Ranger was also killed (as Jay leaving CBM said clearly it is READY to be produced) - CBM management was poor - we all share this opinion - but they are DEAD and you can't turn time back.

[pandy71]

Originally Posted by Cyprian

ok, but the bus (and chipsets) is 3.5MHz 32bit

But it can use DDR in some limited situations...

[Thomas Richter]

Originally Posted by pandy71

PCI killed Tseng...

Well, to the degree that they didn't have a speed advantage anymore. But there is more, actually. Tseng neither had an integrated RAM-DAC, which caused all silly limitations. If you look at the Merlin, it was seriously limited in the horizontal resolution. Even though the RAMDAC could do truecolor, the Tseng could not deliver the necessary amount of data simply because its horizontal timing registers were limited to 2048 samples. 2048/3=682 pixels maximum horizontal solution.

[hammer]

Originally Posted by sandruzzo

Maybe 14mhz ram x 64 bit wide bus?

14 Mhz with 32-bit bus has a theoretical 56 MB/s. 2 clock cycle memory access will reduce this theoretical value to 28 MB/s.

7 Mhz x 32-bit has 28 MB/s.
3.5 Mhz x 32 bit has 14 MB/s.
3.5 Mhz x 16 bit has 7 MB/s.

Only Lisa has a 32-bit custom chip design and has access to 4X memory bandwidth. Lisa's improvements are mismatched with 16-bit Alice.

32-bit Fast RAM is recommended for the 68EC020 CPU.

[hammer]

Originally Posted by pandy71

No one defend Mehdi Ali but you obviously never worked in large company producing mass selling goods... people trying to cut design by using 3..4 cents cheaper part and you suddenly propose 3$ - in large company those 3$ is multiplied immediately by at least 1 million and convince management to accept 3 million $ loss - good luck with this...

Rubbish. You actually defended Mehdi Ali with your personal attack on me.

I'm well aware of Toyota's saving a bolt as a cost reduction tactic. My employer has back-office contracts with vehicle companies' chemical suppliers when they have manufacturing operations in Australia!

Commodore is not competitive in 1993 Gateway 2000's 486SX-33..

When compared to Apple's Quadra 605's 68LC040 for under $1000 USD in Q4 1993, Commodore was gatekeeping 68040 for the elites and asking 486SX-33 level prices for A4000/030 @ 25 Mhz.

There's large price gap between A1200 and A4000/030.

Commodore weakened its 1992 revenue with the A600's sales flop when they listened to Commodore Germany's A300 scope creep and canceled the A500. Irving Gould shouldn't have used Commodore as his personal bank account.

Originally Posted by pandy71

Are you share owner? stock investor with loss on Commodore papers? then sue them...
Most of those you are blaming are probably from very long time decomposed... you could try if you wish to repeat something like this:

What's this topic's purpose?

Originally Posted by pandy71

If you say so - all above chips are for VLB or for PCI bus and usually looks less shiny than competition namely S3 solutions or ATI MACH32.

https://thandor.net/benchmark/32
From Doom benchmarks

ATI MACH32 is not the fastest PCI 2D card i.e. it's a mid-bottom 2D PCI card.

ET6100 2MB PCII(P100) scored 72.59 fps
ATI Mach 32 1MB PCI(P100) scored 61.27fps

Tseng was diminished by 3D and unable to integrate chips as a cost-reduction method at a similar rate as S3 Trio's 1995 release.

ATI purchased Tseng. ATI tends to buy companies with certain advantages e.g. Tseng's strong 2D, ArtX's 3D. ATI is like Commodore's buying habits for tech injection with better management. NVIDIA's in-house RIVA R&D beats ATI.

ET6100's 2D gaming is excellent.

For PCI, I selected S3 after Tseng.

During the S3 ViRGE train wreck, I selected NVIDIA.

Originally Posted by pandy71

I owned W32i with 486DX66 VLB (real DX50 overclocked to 66MHz with IDE on VLB - it was VERY TRICKY to make it work at all) and quickly this fast config become obsolete...

Mainstream 486DX2-66's 33 Mhz front side bus with 33 Mhz VLB exists for a reason.

Pentium P5's mainstream 32-bit PCI has 33 Mhz. 64-bit PCI targeted server and workstations. For graphics, AGP replaced 32-bit PCI.

My point, the PC clone market's advantage is the ability to switch hardware vendors with relative ease. In recent years, NVIDIA's professional software stack is far superior to AMD's Radeon RDNA 1/2/3 professional software stack and it's harder to switch GPU vendors.

Originally Posted by pandy71

PCI killed Tseng...

For Doom
ET6100 2MB PCII (P100) scored 72.59 fps
ATI Mach 32 1MB PCI (P100) scored 61.27fps

Again, Tseng was diminished by 3D and unable to integrate chips as a cost-reduction method at a similar rate as S3 Trio's 1995 release. ATI purchased Tseng.

PC clone market's advantage is the ability to switch hardware vendors with relative ease.

Originally Posted by pandy71

3$ is large difference in mass production when you need to take risk and made pre-orders and this is more than just 3$ as you need to use faster memories or add cache or both... You need to take from pocket some money with risk of no profit... This how life works... In single peaces 3$ are fine, low risk 30..300$ loss well non one like to lose but hey - worst things happen everyday...

There is many ways to speedup Agnus/Allice but they need complex design and may ruin legacy compatibility - i can understand lack of will to take risk...
Ranger was also killed (as Jay leaving CBM said clearly it is READY to be produced) - CBM management was poor - we all share this opinion - but they are DEAD and you can't turn time back.

The profitable Commodore UK was willing to fund an upgraded CD32 e.g. contract some 3rd party Amiga accelerator vendor for a CD32 CPU accelerator in economies of scale. CD32 could be manufactured without the existing 14 Mhz 68EC020.

CD32 powered 68EC020 @ 25 Mhz would step on 486SX-33 priced A4000/EC030's performance range and put Commodore Germany's pricey A4000/EC030 existence in question.
Commodore's 386DX-25 performance range would be in the striking range of CD32 powered 68EC020 @ 25 Mhz, again, the Commodore Germany factor.

No Amiga SKU will step on Commodore's DT486DX2 PC' £760. https://dosdays.co.uk/topics/1993.php there's a price and performance gap between A1200 and A4000/EC030 for Commodore's 486DX-25 placement. The mid-priced A1000+ AGA with 68EC020 @ 25 Mhz SKU is canceled. For 1993, Commodore's best machine for Doom in the mid-price range is the Commodore DT486dx-25 PC.

Corporate politics are factors.

Power Computing's CPU-accelerated Amiga 1200 bundles were too late.