Hombre Amiga Chipset

[Gorf]

Originally Posted by Amigajay

But as a computer with its support i believe it could have got through, a lot of Amiga people jumped ship when they saw the A1200 wasn't the big leap we all hoped for, had Commodore waited as originally intended until 1994 and not rushed out the AA machines, the A500 was still selling well, they should have kept cost reducing that in a new shell.

Both, computer and console, would have been compatible to nothing.
And commodore was never really good at supporting software houses.
No games, no programs.

Then in late 94 released a Hombre A4000 & A1200, the computer would be fast enough not to switch to Windows and fans had the 3D games they desired, heck even 16-bit console games were still selling better than Jaguar/3do etc systems until 95 when the Saturn and PSX hit the west, so the timing was far from being too late.

late 94 would have been way too late for anything. There was no chance to survive on ECS or AGA Amiga so long - they would have needed AAA instead of AGA and even that a little bit earlier.

[grond]

The mess started with ECS not being AGA+chunky. In the same way the next big mistake was doing the ECS-wedge type Amigas (A500+ and A600) in 16 bit instead of 32bit/020 with HDD-Option (basically the A500+ should have been an A1200+chunky). From then on it was only waiting for the end which came a little prematurely due to the losses in the PC business.

[Amigajay]

Originally Posted by Gorf

Both, computer and console, would have been compatible to nothing.
And commodore was never really good at supporting software houses.
No games, no programs.

late 94 would have been way too late for anything. There was no chance to survive on EsC or AGA amiga so long - they would have needed AAA instead of AGA and even that a little bit earlier.

There was talk if i remember from Commodore UK that either ECS or AGA would be on a single chip to bring a certain amount of compatibility to the new machines.

Of course it wouldn't! AGA near enough killed ECS machines dead, they were selling more even after AGA came out, the end result everything got fucked, as i said 16-bit consoles lasted until 95-96, the A500 could have got to the end of 94 easy for a lower price, budget computers are always needed and sell well, PC's were way to high in cost for most familys all through the 80s and 90s.

[Gorf]

Originally Posted by Amigajay

Of course it wouldn't! AGA near enough killed ECS machines dead, they were selling more even after AGA came out, the end result everything got fucked, as i said 16-bit consoles lasted until 95-96, the A500 could have got to the end of 94 easy for a lower price, budget computers are always needed and sell well, PC's were way to high in cost for most familys all through the 80s and 90s.

Consoles: give away the hardware at production cost (or below), earn money with games.
Commodore: earn absolutely nothing with games or software and try to sell cheap hardware.

Well, guess what: consoles are always cheaper and still they earn more money!

cbm could not produce the A500 cheap enough and sales were dropping - thats way they came up with the idea of the the A300 ... and that ended in a even more expensive A600...

Not AGA was killing ECS, the A600 did.

PC's were way to high in cost for most familys all through the 80s and 90s.

True for the 80s but not for the 90s.

And selling weak budgets toys without HD was one of the biggest problems for the Amiga as a platform. Always seen as a toy for kids.

[Amigajay]

Originally Posted by Gorf

Consoles: give away the hardware at production cost (or below), earn money with games.
Commodore: earn absolutely nothing with games or software and try to sell cheap hardware.

Well, guess what: consoles are always cheaper and still they earn more money!

cbm could not produce the A500 cheap enough and sales were dropping - thats way they came up with the idea of the the A300 ... and that ended in a even more expensive A600...

Not AGA was killing ECS, the A600 did.

Consoles are cheaper because they have less parts in them, didn't stop many companies making millions through cheap computers, they sell because they are cheap, smaller profits vs bigger sales = big profits too!

Commodore did the right thing in making a cost reduced A500, they did it wrong by making into an A600, confusing consumers by cutting the numeric keyboard off, imagine if a PC maker did that they wouldn't sell the bloody things! Funnily enough though the A600 outsold the A1200, so it goes to show cheap Amiga's were still wanted.

I still don't agree the A600 killed ECS/A500, A600 sales were strong until Commodore pulled the plug at the end of 92 to try and sell its upcoming A1200, then guess what they didn't have enough inventory, killing sales on both counts, had the A600 had chance to sell on at a cheaper price i have no doubt it could have last a couple more years for Commodore to get Hombre out, A500 sales were over 1 million in 1991 and rising year on year, Commodore couldn't have fucked it up anymore if they tried again!

Anyway, sorry going off topic here.

[Gorf]

Originally Posted by Amigajay

I still don't agree the A600 killed ECS/A500, A600 sales were strong until Commodore pulled the plug at the end of 92 to try and sell its upcoming A1200, then guess what they didn't have enough inventory, killing sales on both counts, had the A600 had chance to sell on at a cheaper price i have no doubt it could have last a couple more years for Commodore to get Hombre out, A500 sales were over 1 million in 1991 and rising year on year, Commodore couldn't have fucked it up anymore if they tried again!

Not true in most markets. Not in UK an Germany and these countries had the largest fanbase.
The A600 was unsellable as soon as the A1200 was announced - loads of them were sent back!

1990 and 1991 was the peak for the A500. but it was clearly a peak and numbers began to drop significantly.

[Hombre40]

Originally Posted by Toffee

Read the AAA specs from DevCon 93 written by Dave Haynie it deffo had no Texture Mapping support. I can upload if you want ?

Just did and indeed you are right, no mention of texture mapping. Also the AAA register list does not mention anything texture mapping related. Yet, I was sure My memory is starting to suck

[grond]

Originally Posted by Gorf

The A600 was unsellable as soon as the A1200 was announced - loads of them were sent back!

The A600s were eventually sold. I bought one new for 299DM (~150EUR) in the summer of 1993, plugged it into the family TV. Two years earlier I would have had to spend five times the money for an A500 plus 1084. There couldn't have been much money to be made with the A600. Commodore obviously was sitting on them and the parts for even more which forced them to sell at a loss reducing short-term losses but canibalising future sales. Hence, the A600 was a big desaster even though it may have sold many units. Nobody explicitly wanted ECS in wedge-type Amigas because there was hardly any benefit over OCS.

AGA just was too little too late. A lot of people that had been saving their money (buying an A600 when you already have an A500 never made that much sense, did it?) spent it on PC hardware. I bought my A1200 used in late 1994 from a guy that was saving for a 486. I guess it hadn't taken him too long to figure out that Doom had more appeal than the same old 2D Amiga games. AGA just never promised to reach a critical threshold. It wasn't the desired advance nor were there enough units sold to make it economically feasible to produce a lot of AGA-only games. This would have been different if AGA had come with the A3000 and had made it to the wedges in the form of a significant update (complete 32bit system) within a short time after the A3000. Commodore was losing the momentum at that time and with the A600 essentially being nothing more than a 1985 computer in a small package it was already too late. Commodore would have kept the Amiga going if they had sold upgrades even before they were desperately needed because the old stuff only made people yawn.

[sandruzzo]

AGA with small changes whould gave some time to put out HOMBRE:

128 kb fast mem
14 mhz- 32 bit blitter and copper
blitter with line mapping
8 16bit audio channel
chunky?( with14mhz -32bit copper would be 12 bit chunky!)

[babsimov]

Originally Posted by Toffee

Yeah AAA had no 3d features.

The built in CPU was twice as fast as PlayStation and was 64 bit vs 32 bit. Hombre had no T + L built in but high end systems could have another HP processor as the main CPU and use the on board chip for doing T + L.

There was no z buffer or texture filtering.

Hombre Blitter and Copper were 64 bit vs AAA 32 bit versions.

Hombre had 12 16 bit audio channels vs AAA 8.

Yeah 4 8 bit playfields all with scrolling and scaling or combined for 32 bit mode.

Hombre was only 2 chips including CPU so would have been much cheaper than AAA systems. They aimed for 2 chip system at sub $40 cost.

These systems would have been fast and cheap !!

Thanx for your answer.

This is probably why they have canceled the AAA for Hombre. If i remember correctly.

[Turrican_3]

Sorry for coming so late to the party...

Today I finally found the time to have a look at those Hombre papers and... wow, so many surprises inside - or at least, stuff I can't remember I had read before:

- AAA still somehow being a thing in 1993-94, but Hombre clearly referred to as the preferred path
- Hombre meeting a mere days before the bankrupcy by CBM guys with... Philips?
- planned "single chip" Motorola CPU + AGA (can't remember the exact one but I don't think it's *that* important) to achieve cheaper hardware
- Hombre actually had nicknames for the chips, in true Amiga fashion
- Hombre would have likely kept at least a little bit of AAA in form of audio (enhanced Mary?)

Definitely a fascinating, yet sad piece of history.

[dhe]

If you ended here, after a google search for the PDF's, you can download here:
https://archive.org/details/Hombre_2...tation_part_1/

[lmimmfn]

Interesting idea, could Commodore have increased the copper so it could have done high res chunky display while leaving the bitplanes/blitter as is? If so what would have been needed for that copper?

[Gorf]

Originally Posted by Ras Voja

Couple of places, but would start with
http://www.bambi-amiga.co.uk/amigahi...gahistory.html

I have no love for hombre just because of CPU chage and no AmigaOS.

However Windows NT would exist, which is crazy? Maybe IBM OS/2 Warp?

I guess nobody at Commodore even asked Microsoft about that!
There was just a rumor MS would port NT to PA-RISC, but that never happened:
https://www.openpa.net/pa-risc_opera...s_history.html

This was just C= grasping at straws in its last weeks, testing the water with unsubstantiated claims.

There were only really three real options:

• HP-UX, the Unix variant from HP itself
• BSD
• porting AmigaOS

all POSIX OSes would have needed some direct rendering interface for the Hombre Chips, which did not yet exist at this time (X11 was is way to slow back then)
Porting AmigaOS ... would probably not been ready in time.

----> total desaster

[Bruce Abbott]

Originally Posted by Gorf

I guess nobody at Commodore even asked Microsoft about that!

Why would they?

There were only really three real options:

• HP-UX, the Unix variant from HP itself
• BSD
• porting AmigaOS

There was really only one option - porting Amiga OS.

But would anyone actually want Amiga OS? Commodore appears to have thought so, which shows how clueless they were. People wanted an OS for sure, but not an incompatible one. And it was the same for the hardware. No matter how amazing Hombre might have been technically, it was useless if it wouldn't run the programs people wanted - ie. PC games and apps.

What surprises me is how detailed the Hombre specs was, right down to IC pin numbering. Someone spent a lot of time working out all that stuff, but nobody stopped to think about what the market for it was (purportedly the chipset covered all bases from games console to high end workstation). How were they going to pitch it to potential buyers when they had no clue themselves?

But sophisticated though it was, you can still see the roots of the original Amiga chipset in there. So again it was designed on the basis that someone other than Commodore's own engineers actually cared about the Amiga's architecture.

All this just shows that Commodore's real mistake was letting the hardware and OS engineers drive the design. Amiga fans would have loved it. They might have sold a few thousand...

With plans like this, we are fortunate that Commodore died when it did.

[Gorf]

Originally Posted by Bruce Abbott

But sophisticated though it was, you can still see the roots of the original Amiga chipset in there. So again it was designed on the basis that someone other than Commodore's own engineers actually cared about the Amiga's architecture.

Yes, in deed.
That is actually the only good part about it - and it is actually more amiga-in-spirit (but also just in theory) than I expected back in the 90s.

All this just shows that Commodore's real mistake was letting the hardware and OS engineers drive the design. Amiga fans would have loved it. They might have sold a few thousand...

Well - as you might expect: here we do not agree.
The Amiga-likeness was not the problem here .... but timing and the fact that is was far from production.

With plans like this, we are fortunate that Commodore died when it did.

unrealistic, too late, laking compatibility ....
but the right direction in terms of vision.

[OlafSch]

@Bruce

I have old amiga magazines where someone from Commodore talked about NT. I am not sure if they (in the company) still knew what was real and what only ideas. Hombre would have been a break compared to the other chipsets and a new processor. Not compatible. So you would have had a revolutionary hardware (perhaps) but without software. Hard to see it successful.

@gorf

Software sells hardware. Nobody buys a hardware without software. The situation in the 90s was diufferent with huge markets like the PC. Spirit even if, the problem would have been to get the content you need to make people wanting to buy it.

[ImmortalA1000]

Hombre, which Dave Haynie claims had better 3D than Sega Saturn, would NEVER have been affordable to put inside a £500 computer or £400 console even if it worked when mass produced. AAA suffered the same fate, it would have been too expensive and inferior to the top end VR based VGA cards for PC-I. These chipsets are only for top-end machines like the proposed A5000.

There is ZERO evidence Commodore could have done anything other than bundle the A1200 with the excellent Apollo 1220/4 accelerator in 1995. There was no magical chipset update for the A1200/CD32 for the £300-400 price bracket. That takes real engineering talent, you know like Jay Miner and his 128 colour Ranger chipset he finished for Commodore by the time of ECS rubbish or R J Mical and Dave Needles hardware sprite scaling Lynx chipset for peanuts and Acorn £1200 RISC PC600 spec computer AND cutting edge 3D visuals into a £600 3DO console. Remember that included the extra profit margin for the 3rd party 3DO manufacturers like Panasonic so in reality if Commodore had not let the original Amiga designers go we could have had a Commodore 3DO spec console for £500 by 1995 easily, even less and sooner if MOS could manufacture the entire thing in house with a licence from Acorn for the ARM CPUs and Jay/RJ/Dave were salaried employees of Commodore not freelance designers.

Commodore after Jack Tramiel == shitshow of a company.

[ImmortalA1000]

Originally Posted by Ras Voja

Really?
Up to 2MB Chip instead 512K Chip and 512K Slow FAST
DblPAL modes incl AHI 16 bit
VGA Productivity
IDE and PCMCIA
Soldered less leaking board

CPU was still weak but heart was strong! And Vampires later came to bite your ass!

Amiga 600 was meant to be the Amiga 300, a £199 32 colour 4 channel sound alternative to the 64 colour 8 channel sound £125 Megadrive.

At £399 when it replaced the A500plus it was more expensive than even the £200-250 Commodore 65 prototype which had 8 channel sound (2 sample, 6 waveforms via dual 8580 SIDS) 256 colour in up to 640x400 AND a blitter with the same sort of CPU power as the SNES really. Another mistake from Commodore, they picked the wrong machine.

The £199 Amiga 300 would have sold well, they just changed the name and doubled the price.

[ImmortalA1000]

Originally Posted by Ras Voja

This part of CBM future is MOST UNKNOWN and speculative, more then C65.

Likes
- New more 3D orinted chipset, kind of Amiga style
- More standardized board, ATX PCI If I get that
- Windows NT! but it was crossplatform and simple pay for m68k version could do

Dislikes
- CPU Change, why not using 68060 as baseline and invest in Freescale continuation, P2 class?
- No AGA compatibility
- Slow deciding, constant changes similar to falldown of A3000 Plus AGA DSP which was way better then A4000 (SCSI, flicker, DSP ...)

Overal 1 and 2 dislike smake it noi Amiga

There is no speculation, Hombre was never intended to be used in any low end A1200/CD32 replacement at the time it was being worked on, it was for A5000 $3000 machine only priced A4000 replacement.

Hombre chipset for Amiga is like 1983 Rainbow chipset for Atari Sierra/Gazza projects that was going nowhere useful. Neither had cutting edge price/performance although very sophisticated, neither would produce a cheap world beating revolutionary chipset for consoles/home computers like Atari 800/5200 (in Hombre's case £300-400 price point of A1200/CD32 equivalents). A bunch of mediocre engineers taking too long to make a chipset that would be too expensive for home gaming. FAIL. ditto with Hombre and AAA.

Atari's AMY chip was very cutting edge and revolutionary (additive not subtractive synthesis), wiped the floor even with 1990s Creative AWE32 a decade after it BUT it was not commercially viable in 1985 which is why the Atari ST has that crap YM chip. No decent engineering talent at Atari was a big problem (or Commodore in the time of 520ST).

Commodore would never make money and save their finances with an Amiga 5000, the money was in A1200/CD32 replacement home gaming mass market OR budget Windows x86 home/family High Street PC compatibles.

CD32/A1200 machines were "7mhz" crippled 14mhz 020 with a 256 colour mode slower than Amiga 1000 extra half brite due to shitty DMA bus design and exactly the same sound hardware....wooopee doo.

Anybody can make a 'better than Sega Saturn' 3D chipset, making it as cheap as a Sega Saturn chipset is where the talent comes in.

Nolan Bushnell will tell you the same thing, a shit salesman wants a Cadillac for the price of a Ford, Commodore engineers were trying for N64 chipset level but Commodore to make money needed it to be for the price of a 3DO chipset. NEVER going to happen. Hombre is another mediocre attempt from Dave Haynie and pals that would never work in the mass market for the desperately needed A1200/CD32 replacement Commodore A4000/040 cost DOUBLE the price of the identical spec Macintosh LC475.

R J Mical and Dave Needle did a good job for 3DO, Commodore engineers in the 1990s were beneath those two engineers they forced out after buying their company.

Even if Hombre existed assholes like US Gold/Ocean would made the same old shit quality arcade conversions so it doesn't even matter. Shitty conversions is what killed A1200/CD32 anyway. As the Saturn adverts famously said 'reality always hurts'