@bhabbott

Quote:

bhabbott wrote: @agami Avoid 1994 bankruptcy perhaps, but not much more.

Oh I agree.
Any action in the realm of our reality which would've resulted in the avoidance of the Commodore bankruptcy in 1994 would ultimately only be kicking the proverbial can just a little down the road.
But try and explain that to a person who lives in their own constructed reality.

Everything you've written is very reasonable. I was just trying to illustrate how @ppcamiga1's mind works with his:
10 Phase 1 - Collect underpants
20 Phase 2 - GOTO 30
30 Phase 3 - Profit
hindsight view of the world.

_________________
All the way, with 68k

@bhabbott

Quote:

ppcamiga1 wrote: doom works well on atari falcon little faster than amiga 1200. so doom in low detail is possible on amiga 1200 with fast ram. commodore bankrupt because aga has not chunky pixels.

See what I mean?

Last edited by agami on 12-Dec-2023 at 01:37 AM.

_________________
All the way, with 68k

@cdimauro

Quote:

cdimauro wrote: @bhabbott: AGA was too little even on 1990.

Right. It's not worth discussing the subject with people eho make such statements. Bye.

@cdimauro

Quote:

FACT is: Amiga was wide spread due to its LOW-END, GAMING machines. .

FACT: Amiga OCS started from a mid-high SKU with the Amiga 1000 in 1985. Commodore applied their cost reduction skills to the Amiga 1000 design and created a cost-reduced 1987 Amiga 500 ($699 USD).

A similar pattern is applied for Amiga 1200 which is largely based on AA3000+.

Amiga Ranger was largely completed in 1987 and its R&D evolution should have continued for the 1990 release.

The problem is Commodore management's time wasting and misdirection.

PC cloners played the same cost-reduced approach for IBM's VGA and 8514.

Quote:

That should have been the PRIMARY focus of Commodore (managers AND engineers). You can ALSO think about providing MID/HIGH-end systems IF YOU CAN, but the focus should have been the above low-end AND, specifically, the gaming market.

Higher-cost technology should exist for applying cost-reduction methods.

Sun GX quadrilateral 3D and Lockheed Martin's Real3D texture mapping workstation class technologies influenced Sega's Saturn design.

Nintendo 64 is directly influenced by SGI workstation OpenGL with the SGI team directly working on the N64's chipset. This SGI game console team's personnel created ArtX company which is purchased by ATI.

Commodore's Amiga Hombre chipset has been influenced by HP's PA-RISC workstation CPUs.

Game console vendors are raiding workstation graphics and designing cost-reduction products. The same is true for current generation game consoles based on PC's gaming GPU flagships.

PS5 Pro's November 2024 GPU is based on AMD's RX 7800 XT IP with full hardware BVH transversal improvements. Today's near-flagship PC GPU is tomorrow's game console.

Amiga OCS's 4096 color palette design target wasn't an accident e.g. NEC's 1981 PC-98's 4096 color palette.

Quote:

For this reason, a 486 with a phenomenal SVGA, a Mac-II and even the Sharp X68000 are completely irrelevant as a reference for such market.

Wrong. Commodore engineers should have the chance to master high-performance graphics IP before delivering a cost-reduced version. Commodore management has to provide the necessary leadership and funds for this R&D process.

Again, game console vendors are raiding workstation graphics and designing cost-reduction products.

Last edited by Hammer on 12-Dec-2023 at 02:32 AM.

_________________
Ryzen 9 7900X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB
Amiga 1200 (Rev 1D1, KS 3.2, PiStorm32lite/RPi 4B 4GB/Emu68)
Amiga 500 (Rev 6A, KS 3.2, PiStorm/RPi 3a/Emu68)

@bhabbott

Quote:

bhabbott wrote: Avoid 1994 bankruptcy perhaps, but not much more. 3rd quarter 1993 would be too late. The Quadra 605 is a red herring.

1. Quadra 605's 68LC040 price is competitive against PC clones 486SX-25/486SX-33.

Provide under $1000 USD Amiga 1200 with 68LC040 accelerator solution in 1993.

Quadra 605 example shows Motorola's 68LC040 wasn't a problem during 1993.

2. Apple's PowerPC wasn't ready during 1993. The goal is to keep Apple's customers within Apple's ecosystem during H2 1993.

3. Quadra 605 is PowerPC upgradable.

4. MacOS has matured QuickDraw (RTG) middleware infrastructure, hence many different Mac models weren't a large issue.

5. Quadra 605 design shares with LC 475 and Performa 475 variants. Performa 475 was discontinued on April 1, 1996.

Making a big deal about Quadra 605's October 1994 EOL shows your ignorance.

1993 Amiga was squeezed from the bottom by SNES and the top via failing 386DX-33/486SX PC prices while Apple avoided being squeezed out of the market.

Apple picked a market and stayed relevant.

Many Amiga owners jumped ship in 1993.


Quote:

Firstly it was useless without the companion monitor, which wasn't cheap.

The Commodore 1942 monitor wasn't cheap either.

Quote:

And then it only ran Mac software.

I was running Mac 68K software on my Amiga 3000 e.g. Microsoft Office 4.2.

Microsoft Office's market is larger than Amiga's Video Toaster 4000 market.

Quote:

That model only sold for 1 year - I wonder why?

1. Quadra 605 design is common with LC 475 and Performa 475 variants. Performa 475 was discontinued on April 1, 1996.

Making a big deal about Quadra 605's October 1994 EOL shows your ignorance.

2. Apple's Quadra 605 with 68LC040 was the Q4 1993 XMas stop-gap measure to mitigate against the 486SX-33 PC clone's aggressive pricing.

The goal is to keep Apple's customers within Apple's ecosystem since Apple's PowerPC wasn't ready in 1993.

Again, Apple picked a market and stayed relevant.

Not many families can sustain two 32-bit computer platforms with 386DX-33 and 68030 Amiga personal computers at the same time. My school friends shifted from A500 to 486 PC build in 1993, hence why I have another A500 Rev5 after my Dad sold our A500 Rev 6.

From A500 Rev5's components, I rebuilt the A500 Rev6 in 2020.


Quote:

Chunky graphics in 640x480 256 colors is not much use by itself. You also need a fast CPU and faster blitter etc. This is no simple upgrade to OCS, and the machine would be expensive. But people wouldn't buy it at the same or higher price as a PC, which makes competing head-to-head with PCs very difficult when clone makers were churning out stuff cheaper than any name-brand manufacturer could.

My major 1993 example is Gateway 2000 PC brand and Apple's Quadra 605.
My 386DX-33 PC clone is a local Sydney build with a Taiwanese motherboard.

Amiga 1200's AGA didn't maximize 32-bit Chip RAM since graphics hardware wasn't a full 14 Mhz 32-bit chipset.



Quote:

The only reason Commodore could make the A500 so cheap was having their own chipset in a minimalist design.

In 1992, Commodore's chip integration was behind the competition.

1992 era 386DX motherboard with high integration via SIS's super IO chipset.

it's an expandable desktop PC with a low chip count.

Amiga OCS/ECS was an aging chipset in 1991. SNES arrived in Amiga's core European markets in 1992 along with Commodore's AGA.

SNES crushed lower-cost Amiga gaming use case.

Stock A1200 performs like an enhanced 16-bit computer instead of delivering new "32-bit gaming" experiences i.e. it performs like 386SX-16 with a reasonable VGA-like solution (minus chunky 8-bit pixel).

Adding 32-bit Fast RAM helps with AGA's framebuffer performance.

Quote:

640x480 8 bit was overkill. 320x200/256 8 bit was the sweet spot for games - AGA would have been OK and perfect with chunky pixels at that resolution. But it needed to work at 15kHz so people could use their TV or 1084 monitor. It still wouldn't be much use without a faster CPU though. And why do you need it anyway? Only to make porting PC games easier. That means the Amiga would still be playing catchup while not being IBM compatible. For the Amiga to stay relevant it had to continue carving a niche between consoles and PCs. The A1200 did that - it was just a little too late.

Amiga 1200's AGA didn't maximize 32-bit Chip RAM since the 16-bit Blitter is wasting its 32-bit bus cycle slot.

EA's uncompleted Magic Carpet AGA build has no ground textures and it's a dead duck.

A1200 didn't deliver a "32-bit" gaming experience.

In your extract, where's "ECS A1000 Jr" time-wasting?

September 1989 for starting AA R&D is too late for A3000.

AA3000+ design cost reduced for A4000 e.g. replace SCSI for PIO Mode 0 IDE, remove AT&T DSP3210. A4000's construction is expensive compared to SuperI/O equipped 386DX/486DX motherboards.


https://youtu.be/BaTjwo1ywcI
Dave Haynie's The Deathbed Vigil

https://youtu.be/BaTjwo1ywcI?t=419

New Engineering management took over in Summer 1991. By October, all "AA" projects had been cancelled. Only prototype A3000+ systems were built for OS development.
The A300, a super cheap Amiga was remade into the A600, an A500 replacement. More costly than the A500, it offered users less. The A500 was the first still popular machine in Commodore history to be cancelled


https://youtu.be/BaTjwo1ywcI?t=510

By April 1992, there is still no "AA" machine. A mid-range ECS-based system is mandated by management, built and rejected by all Commodore sales divisions
Finally, AA systems can be made. The high-end A4000 is a slap-together of A3000 parts and ideas from the mid-range system.


https://youtu.be/BaTjwo1ywcI?t=822
AAA prototype.

Last edited by Hammer on 12-Dec-2023 at 05:52 AM.
Last edited by Hammer on 12-Dec-2023 at 05:51 AM.
Last edited by Hammer on 12-Dec-2023 at 04:58 AM.
Last edited by Hammer on 12-Dec-2023 at 04:53 AM.
Last edited by Hammer on 12-Dec-2023 at 04:49 AM.
Last edited by Hammer on 12-Dec-2023 at 04:37 AM.
Last edited by Hammer on 12-Dec-2023 at 04:29 AM.

_________________
Ryzen 9 7900X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB
Amiga 1200 (Rev 1D1, KS 3.2, PiStorm32lite/RPi 4B 4GB/Emu68)
Amiga 500 (Rev 6A, KS 3.2, PiStorm/RPi 3a/Emu68)

@ppcamiga1

Quote:

ppcamiga1 wrote: doom works well on atari falcon little faster than amiga 1200. so doom in low detail is possible on amiga 1200 with fast ram. commodore bankrupt because aga has not chunky pixels.

https://youtu.be/W6utQR-tKt8?t=1179
Amiga 1200 with Fast RAM plays Doom like 386SX-16 with IBM VGA.

Certain A1200 batches have timing bugs with expansion cards like my A1200 rev 1D4. A timing fix was applied for TF1260.

_________________
Ryzen 9 7900X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB
Amiga 1200 (Rev 1D1, KS 3.2, PiStorm32lite/RPi 4B 4GB/Emu68)
Amiga 500 (Rev 6A, KS 3.2, PiStorm/RPi 3a/Emu68)

@bhabbott

Quote:

bhabbott wrote: @cdimauro Quote: cdimauro wrote: @bhabbott: AGA was too little even on 1990. Right. It's not worth discussing the subject with people eho make such statements. Bye.

I know, I know: blind fanaticals aren't able to accept anything which they don't like, reality included.

You can go as well, join other lotus eaters and be happy...


@Hammer

Quote:

Hammer wrote: @cdimauro Quote: Nobody had the crystal ball to see on 1989 and 1990 what could have happened on 1993. Please stop this incorrect representation of the facts, because it's historically inaccurate. That's a false narrative. You didn't factor in Moore's law as the constant drumbeat pace.

Hey, I worked at Intel! But not even there we had crystal balls to look at the future evolution of games...
Quote:

PC VGA and PC chipset cloners followed Moore's law with highly integrated ASICs. For the 1990 product, the A3000's chip integration and high chip count are inferior. https://vintageapple.org/pcworld/pdf/PC_World_9011_November_1990.pdf Page 354 of 364 for the US market November 1990. Full 32-bit 386 CPU @ 20Mhz PC with Phoenix BIOS, 1 MB RAM, PSU, 101 Keyboard, 1.44 FDD, MS-DOS 3.3, IDE controller, and IDE Seagate The base system asking price is $999. VGA card with 48 MB HDD, extra $850 SVGA card with 20 MB HDD extra $709 SVGA card with 48 MB HDD extra $1019 A low-cost VGA color monitor can be obtained.

Now look at how much an Amiga 500 costed, which have NOT required a monitor but usually was attached to a TV: is the above comparabile with this very low-end GAMING machine?!?

That's the point of the whole discussion (and of my articles). Do you understand it now? Because I've no other ways to let you know it.
Quote:

Direct competition is the main driver for pushing PC's lower asking prices, hence the direction is set for the 1993 situation.

Sure, thanks to your crystal ball...
Quote:

https://archive.org/details/eu_BYTE-1990-05_OCR/page/n209/mode/1up Commodore Amiga 3000 review by Byte Magazine, May 1990 The reviewer correctly noted that the lack of native 8 and 24-bit graphics modes is a concern and that Commodore needed to address this to stay competitive. Commodore's asking price for A3000 @ 16Mhz 68030 is $3495 and 25 Mhz version for $3995. The mentioned Amiga 3000 has 2 MB memory (1 MB Fast, 1 MB Chip) and 50 MB SCSI HDD. On the subject lack of native 8 and 24-bit graphics modes, Commodore's Amiga Product Manager Keith Masavage commented Commodore has both short-term and long-term solutions are the works. (read Byte Magazine's review). Recall that the AA3000+ AGA chipset was completed in Feb 1991. Commodore management wasted time on ECS-based "A1000+ Jr". ---- Gaming PC 1990 Wing Commander VGA was the killer game app. My Dad noticed certain PC games in 1990 and made plans for a later 386DX-33 PC clone build in early 1992. In 1992, My Dad obtained our ex-corporate A3000/030 (25Mhz, KS 2.04, 1MB Fast, 1 MB Chip, Amiga Vision) from corporate contacts which is not a standard retail path. A500 Rev 6 (1MB RAM) was sold at a similar time to help fund the ex-corporate A3000/030 purchase.

How much costed the 386DX33 PC, the Amiga 3000/030 and the A500?

After that, you can do the math.
Quote:

The signs for 1993 already exist in the 1990's PC price decline direction. Quote: Parts of Commodore were already aware of the 256-color VGA problem in 1990! Commodore sales teams did NOT purchase "A1000 Jr".

Do you understand that 1990 is the year were the ECS chipset was DELIVERED? So, they COMPLETED the development (which started BEFORE) and then sold the machines.

If something happened on the same year, then... it was already too late for the ECS release.

Do you think that Commodore engineers had a time machine which allowed them to go back and fix / enhance ECS only because Wing Commander was released?!?

"Oh, look: there's Wing Commander for PC now! Let's go back when we're working on the drawing board, then we can add the packed/chunky graphics and be ready for games like that.

Quote:

Hammer wrote: @cdimauro Quote: FACT is: Amiga was wide spread due to its LOW-END, GAMING machines. . FACT: Amiga OCS started from a mid-high SKU with the Amiga 1000 in 1985. Commodore applied their cost reduction skills to the Amiga 1000 design and created a cost-reduced 1987 Amiga 500 ($699 USD). A similar pattern is applied for Amiga 1200 which is largely based on AA3000+.

How much they've costed compared to PCs and high-end machines like Mac-II (and successors) and X68000?

This is the point. Do you understand it?
Quote:

Amiga Ranger was largely completed in 1987

That's simply false: another invention...
Quote:

and its R&D evolution should have continued for the 1990 release.

It was expensive.
Quote:

The problem is Commodore management's time wasting and misdirection.

Same with engineers, which had no vision of what was really needed for the PRIMARY market: games.
Quote:

PC cloners played the same cost-reduced approach for IBM's VGA and 8514.

Sure, but it doesn't change the situation: those were very expensive machines.

Whereas the Amiga market was the low-end/cost gaming...
Quote:

Quote: That should have been the PRIMARY focus of Commodore (managers AND engineers). You can ALSO think about providing MID/HIGH-end systems IF YOU CAN, but the focus should have been the above low-end AND, specifically, the gaming market. Higher-cost technology should exist for applying cost-reduction methods.

"No new chips..."

Have you forgot it?
Quote:

Sun GX quadrilateral 3D and Lockheed Martin's Real3D texture mapping workstation class technologies influenced Sega's Saturn design. Nintendo 64 is directly influenced by SGI workstation OpenGL with the SGI team directly working on the N64's chipset. This SGI game console team's personnel created ArtX company which is purchased by ATI. Commodore's Amiga Hombre chipset has been influenced by HP's PA-RISC workstation CPUs. Game console vendors are raiding workstation graphics and designing cost-reduction products. The same is true for current generation game consoles based on PC's gaming GPU flagships. PS5 Pro's November 2024 GPU is based on AMD's RX 7800 XT IP with full hardware BVH transversal improvements. Today's near-flagship PC GPU is tomorrow's game console.

That's garbage. The context was 1990!
Quote:

Amiga OCS's 4096 color palette design target wasn't an accident e.g. NEC's 1981 PC-98's 4096 color palette.

Even the PGA, but it was a BIT expensive to be used for games...
Quote:

Quote: For this reason, a 486 with a phenomenal SVGA, a Mac-II and even the Sharp X68000 are completely irrelevant as a reference for such market. Wrong. Commodore engineers should have the chance to master high-performance graphics IP before delivering a cost-reduced version. Commodore management has to provide the necessary leadership and funds for this R&D process.

They have both their responsibilities.

Don't remove the engineers from the equation: they never understood what was really needed for the Amiga, as it can be clearly seen (same Blitter, same Paula).
Quote:

Again, game console vendors are raiding workstation graphics and designing cost-reduction products.

Have you checked how much were costing the consoles of the time?

Just to stay A BIT on the topic / context...
Quote:

Hammer wrote: @bhabbott Quote: bhabbott wrote: Avoid 1994 bankruptcy perhaps, but not much more. 3rd quarter 1993 would be too late. The Quadra 605 is a red herring. 1. Quadra 605's 68LC040 price is competitive against PC clones 486SX-25/486SX-33. Provide under $1000 USD Amiga 1200 with 68LC040 accelerator solution in 1993. Quadra 605 example shows Motorola's 68LC040 wasn't a problem during 1993. 2. Apple's PowerPC wasn't ready during 1993. The goal is to keep Apple's customers within Apple's ecosystem during H2 1993. 3. Quadra 605 is PowerPC upgradable. 4. MacOS has matured QuickDraw (RTG) middleware infrastructure, hence many different Mac models weren't a large issue. 5. Quadra 605 design shares with LC 475 and Performa 475 variants. Performa 475 was discontinued on April 1, 1996. Making a big deal about Quadra 605's October 1994 EOL shows your ignorance. 1993 Amiga was squeezed from the bottom by SNES and the top via failing 386DX-33/486SX PC prices while Apple avoided being squeezed out of the market.

And, just by any change, how much costed a SNES and the given PC?

"Just" to see how they compared with Commodore's low-end machines...
Quote:

Apple picked a market and stayed relevant.

But you do NOT (want to?) understand the same elementary concept for Commodore and its game market...
Quote:

Many Amiga owners jumped ship in 1993.

If they had a lot money... why not.
Quote:

Quote: Firstly it was useless without the companion monitor, which wasn't cheap. The Commodore 1942 monitor wasn't cheap either.

But it was NOT required for the low-end machines: attaching them to the TVs was enough.

I know that you were reach with you daddy which bought a lot of cool stuff but, hey, you were exceptions and NOT the norm for the Amiga market...
Quote:

Quote: And then it only ran Mac software. I was running Mac 68K software on my Amiga 3000 e.g. Microsoft Office 4.2. Microsoft Office's market is larger than Amiga's Video Toaster 4000 market.

Ah, yes: the typical things owned / used by the typical Amiga users...
Quote:

Amiga OCS/ECS was an aging chipset in 1991. SNES arrived in Amiga's core European markets in 1992 along with Commodore's AGA. SNES crushed lower-cost Amiga gaming use case.

What do you expect since the maximum was represented by the crappy ECS?

But no, for you it was important to target the high-end market, eh! 32-bit Chip RAM running at 25Mhz and a 32 bit processors: exactly the same stuff which was used by a SNES, right?
Quote:

Stock A1200 performs like an enhanced 16-bit computer instead of delivering new "32-bit gaming" experiences i.e. it performs like 386SX-16 with a reasonable VGA-like solution (minus chunky 8-bit pixel). Adding 32-bit Fast RAM helps with AGA's framebuffer performance.

Expensive. And too late: AGA wasn't competitive even on 1990...
Quote:

Quote: 640x480 8 bit was overkill. 320x200/256 8 bit was the sweet spot for games - AGA would have been OK and perfect with chunky pixels at that resolution. But it needed to work at 15kHz so people could use their TV or 1084 monitor. It still wouldn't be much use without a faster CPU though. And why do you need it anyway? Only to make porting PC games easier. That means the Amiga would still be playing catchup while not being IBM compatible. For the Amiga to stay relevant it had to continue carving a niche between consoles and PCs. The A1200 did that - it was just a little too late. Amiga 1200's AGA didn't maximize 32-bit Chip RAM since the 16-bit Blitter is wasting its 32-bit bus cycle slot.

And the same with the processor, which wasn't able to take all bus cycles which it could have had...

OK, rest is the usual garbage crossing years and mixing all together, as it usually happens with you.

P.S. No time to read again. And not important...

Last edited by cdimauro on 12-Dec-2023 at 10:13 PM.

@cdimauro

I agree mostly with your article.
On chipsets, AGA was far too little at a time when it became too late to consider it.
Even Dave.Haynie has said it should have been AAA however, due to certain decisions, it was AGA as a stop-gap. This was the beginning of the end. I saw this at the time AGA was announced instead of AAA.
There were just too many bottlenecks in AGA.
AAA was a design that was there to address these shortcomings that the engineers knew were holding the machine back, and they knew it very well. Dave was one who was quite frustrated, even angry by the decision.

i remeber attending an AMIGA meeting at the Universtiy here in Australia and the marvels of AAA and that the delay will be a little longer to iron out some issues, only to read it was replaced by AGA in a magazine later on. I remeber that sinking feeling in my gut and thinking, that was a stupid mistake. AGA didn't seem a lot better with the same old bottlenecks.

Last edited by SHADES on 12-Dec-2023 at 09:15 PM.
Last edited by SHADES on 12-Dec-2023 at 09:12 PM.
Last edited by SHADES on 12-Dec-2023 at 09:07 PM.

_________________
It's not the question that's the problem, it's the problem that's the question.

@SHADES: indeed. Same delusions.

However my A1200 was cheap and affordable, at least. I couldn't had the chance to take something much more expensive.

@cdimauro

Quote:

Hey, I worked at Intel! But not even there we had crystal balls to look at the future evolution of games...

1. Constant PC hardware evolution helps with lowering the prices for older PC hardware.

Chipset integration helps with lowering motherboard costs. Intel released the 420TX chipset in 1992.

Intel 82395DX cache controller for 386 wasn't cheap i.e. 196-pin PQFP for USD $90 and $109 for 25- and 33-MHZ versions in quantities of 1000 respectively.

I purposely cited the 386DX-era motherboard from ASUS with SIS chipset and on-board cache as a competitive example. The picture shows a 128 KB cache setting.

There are several 386-era chipsets from C&T, UMC, VIA, SIS, and 'etc'.

2. X86 CPU clones have downward price pressures.

3. Intel is aware of the RISC threat such as from 1991 Advanced Computing Environment (ACE).

Quote:

Now look at how much an Amiga 500 costed, which have NOT required a monitor but usually was attached to a TV: is the above comparabile with this very low-end GAMING machine?!? That's the point of the whole discussion (and of my articles). Do you understand it now? Because I've no other way to let you know it.

My 1989 Amiga 500 Rev6A comes with a Commodore 1084S monitor, the cost is about AUD $1400. In 1990, a 512 KB expansion card was fitted.

In 1992, the spending pattern was repeated for 386DX-33/ET4000X PC clone with IBM PS2 Model 55SX trade-in. 1-year delay into late 1993 would have landed with 486SX.

My dad purchased an ex-corporate A3000 (25Mhz Kickstart 2.04, 1MB Chip, 1 MB Fast) and sold the mentioned A500 in early 1992. The cost about $900. The 1084S monitor was kept.

SNES arrived in Amiga's core European market in 1992 and it covered low-priced gaming.

AUD is usually 30% weaker than USD.

Quote:

Sure, thanks to your crystal ball...

PC's declining price pattern was noticeable in 1990 when 386DX-20-based PC clones were entering my family's annual computer hardware budget range.

Quote:

How much costed the 386DX33 PC, the Amiga 3000/030 and the A500?

1. My 1989 Amiga 500 Rev6A comes with a Commodore 1084S monitor, the cost is about AUD $1400. In 1990, a 512 KB expansion card was fitted.

2. My dad purchased an ex-corporate A3000 (25Mhz Kickstart 2.04, 1MB Chip, 1 MB Fast) and sold the mentioned A500 in early 1992. The cost about $900.

My family's annual computer hardware budget is around $1500 AUD. If this budget is not spent, it goes to a house mortgage.

June 1992 in Australia



Quote:

Do you understand that 1990 is the year were the ECS chipset was DELIVERED? So, they COMPLETED the development (which started BEFORE) and then sold the machines.

Do you understand that 1989 era Amiga 500 Rev 6A has ECS Agnus?

Except for 2MB Chip RAM addressing, 1989 era Amiga 500 Rev 6A's ECS Agnus 8372A has all the improvements like A3000's 8372AB.

1989 era Amiga 500 Rev 6 PCB design has 2MB Chip RAM jumpers and supports ECS Agnus 8372AB or 8372B.

Product management reserved 2MB Chip RAM for the A3000.

A3000's ECS Denise's four-color productivity modes are useless for games and inferior to VGA's 640x480p 16 colors.

A3000's Amber with framebuffer is a workaround for flicker-free 640x400p 16 colors on Amiga chipset and PC VGA (31.5 kHz, 60 hz) monitor.


Quote:

If something happened on the same year, then... it was already too late for the ECS release. Do you think that Commodore engineers had a time machine which allowed them to go back and fix / enhance ECS only because Wing Commander was released?!?

Do you understand that the 1989 era Amiga 500 Rev 6 PCB design has 2MB Chip RAM jumpers?

Do you understand that 1989 era Amiga 500 Rev 6A's ECS Agnus 8372A has all the improvements like A3000's 8372AB except 2MB Chip RAM addressing?

Quote:

"Oh, look: there's Wing Commander for PC now! Let's go back when we're working on the drawing board, then we can add the packed/chunky graphics and be ready for games like that.

Amiga 2500UX's Amix has support for TIGA based A2410. The A2410 card is directly supported by the Amiga Unix SVR4 X11R4 from the X-Windows R4.

Commodore's A2410 TIGA did nothing for Amiga chipset evolution.

Quote:

How much they've costed compared to PCs and high-end machines like Mac-II (and successors) and X68000? This is the point. Do you understand it?

Mastery of certain technology is equally important.


Quote:

That's simply false: another invention.

https://en.wikipedia.org/wiki/Amiga_Ranger_Chipset

Jay Miner made sure that the Ranger chipset was completed and fully tested before he left Commodore in hope that one day the company would release it. But at that time, VRAM was considered expensive compared to DRAM, so Commodore refused to release Ranger for its high price which was unsuitable for the low-end systems like Amiga 500 to cover its costs. Dale Luck is in possession of a Ranger system prototype, but it's unknown what state it is in.



https://web.archive.org/web/20090216124135/http://amigau.com/aig/comment-6.html

Q: Will there ever be an advanced version of the chipset?

A: Well, Jay Miner isn't working on anything right now... [RUMOR ALERT] The chip folks left in Los Gatos who are losing their lease in March were at one time thinking about 1k square 2meg chip space 128-color graphics, although still with 4 bit colour DACs though... and even stuff like a blitter per plane (!!) They were supposed to be done now, in the original plans; the chip designers will be gone in March, but the design may (?) continue in West Chester. Maybe they'll be here two years from now.

Q: What will happen to the unused Los Gatos A2000 design?

A: ??????


https://web.archive.org/web/20030506234259/http://amigau.com/aig/jayinterview2.jpg
Jay Miner Interview

Jay Miner: In spite of very limited manpower, we managed to finish the 1.2 software release; and design a revised set of custom chips for the next generation of Amiga computers
....

They increase the display address range to two megabytes. These chips are completed and tested and only require a computer and memory to hold them together.



1989 era Amiga 500 Rev 6A has 2 MB Chip RAM jumpers for 2MB ECS Agnus.

Quote:

It was expensive.

DRAM's performance also improves over time.

ET4000AX supports either 16-bit VRAM or 32-bit DRAM.

SDRAM appeared in 1992 and SGRAM in 1994.

Quote:

Same with engineers, which had no vision of what was really needed for the PRIMARY market: games.

The difference between 3DO (quadrilateral polygon 3D) vs Sony's Kutaragi-led PlayStation (triangle polygon 3D) teams.

3DO had a $699 price in 1993 which is similar to A500's $699 in 1987, but 3DO is not a general-purpose computer.

Quote:

That's garbage. The context was 1990!

https://old.hotchips.org/wp-content/uploads/hc_archives/hc01/3_Tue/HC1.S8/HC1.8.2.pdf
Sun GX Series Graphics Workstations in 1989. One of its key personnel co-founded NVIDIA along with an AMD microprocessor designer who is the current CEO.

Curtis Priem co-founded NVIDIA with a history from IBM PGA (released in 1985, 4096 color palette, 256 colors) and Sun GX. He retired from NVIDIA in 2003.

NVIDIA's NV1 was released in 1995. Market interest in the product quickly ended when Microsoft announced the DirectX specifications, based upon triangle polygon rendering.

NVIDIA's intellectual foundation was from 1985 IBM PGA and late 1980s Sun GX.

Sun's TEC features
Common interface with SPARC, 680x0 and 80x86
51 MFLOPS single precision for transforms. Supports scaling, rotation,
Integer, signed fixed point, binary, and IEEE single precision floating point.


Meanwhile, Dave Haynie's AA3000+'s DSP3210 has 33 MFLOPS SP FP.


Quote:

Even the PGA, but it was a BIT expensive to be used for games...

Curtis Priem designed the 1985-era IBM PGA and it was expensive but it set the foundations for IBM's 256-color VGA in 1987 with many cost-reduced VGA cloners following this IBM standard.


Quote:

They have both their responsibilities. Don't remove the engineers from the equation: they never understood what was really needed for the Amiga, as it can be clearly seen (same Blitter, same Paula).

Sony's CEO listens to Kutaragi. Many of Kutaragi's meetings are with hardware prices and trade-offs.

Last edited by Hammer on 13-Dec-2023 at 08:09 AM.
Last edited by Hammer on 13-Dec-2023 at 08:02 AM.
Last edited by Hammer on 13-Dec-2023 at 07:52 AM.
Last edited by Hammer on 13-Dec-2023 at 07:37 AM.
Last edited by Hammer on 13-Dec-2023 at 07:34 AM.
Last edited by Hammer on 13-Dec-2023 at 07:09 AM.
Last edited by Hammer on 13-Dec-2023 at 07:01 AM.
Last edited by Hammer on 13-Dec-2023 at 07:00 AM.

_________________
Ryzen 9 7900X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB
Amiga 1200 (Rev 1D1, KS 3.2, PiStorm32lite/RPi 4B 4GB/Emu68)
Amiga 500 (Rev 6A, KS 3.2, PiStorm/RPi 3a/Emu68)

@Hammer

Quote:

Hammer wrote: @cdimauro Quote: Hey, I worked at Intel! But not even there we had crystal balls to look at the future evolution of games... 1. Constant PC hardware evolution helps with lowering the prices for older PC hardware. Chipset integration helps with lowering motherboard costs. Intel released the 420TX chipset in 1992. Intel 82395DX cache controller for 386 wasn't cheap i.e. 196-pin PQFP for USD $90 and $109 for 25- and 33-MHZ versions in quantities of 1000 respectively. I purposely cited the 386DX-era motherboard from ASUS with SIS chipset and on-board cache as a competitive example. The picture shows a 128 KB cache setting. There are several 386-era chipsets from C&T, UMC, VIA, SIS, and 'etc'.

Which does NOT mean that we know what should have happened with the videogames in some years.

Simple, elementary, logic...
Quote:

2. X86 CPU clones have downward price pressures.

Same as above.
Quote:

3. Intel is aware of the RISC threat such as from 1991 Advanced Computing Environment (ACE).

But it wasn't aware of what could have happened in the future regarding videogames.

Otherwise it wouldn't have spent BILLIONS on the Larrabee project, just to nominate another bug project which failed due to missing crystal balls...
Quote:

Quote: Now look at how much an Amiga 500 costed, which have NOT required a monitor but usually was attached to a TV: is the above comparabile with this very low-end GAMING machine?!? That's the point of the whole discussion (and of my articles). Do you understand it now? Because I've no other way to let you know it. My 1989 Amiga 500 Rev6A comes with a Commodore 1084S monitor, the cost is about AUD $1400. In 1990, a 512 KB expansion card was fitted. In 1992, the spending pattern was repeated for 386DX-33/ET4000X PC clone with IBM PS2 Model 55SX trade-in. 1-year delay into late 1993 would have landed with 486SX. My dad purchased an ex-corporate A3000 (25Mhz Kickstart 2.04, 1MB Chip, 1 MB Fast) and sold the mentioned A500 in early 1992. The cost about $900. The 1084S monitor was kept.

Hey, we know that you were rich and your daddy bought you several EXPENSIVE toys.

But this was NOT the case for the average Amigan, which was able to afford the low-entry machines (yes: the ones which were most sold by Commodore).

Have you got the concept or not? I don't know how else should I tell you...
Quote:

SNES arrived in Amiga's core European market in 1992 and it covered low-priced gaming.

TWO years after the ECS and using... could you please tell me which kind of CPU, video and audio processors it was using on 1992?

Just to see if you finally understand what means having a low-cost machine running games...
Quote:

Sure, thanks to your crystal ball...

PC's declining price pattern was noticeable in 1990 when 386DX-20-based PC clones were entering my family's annual computer hardware budget range.[/quote]
Which was still NOT the target of the people that were buying the low-cost Amiga models for gaming...
Quote:

Quote: How much costed the 386DX33 PC, the Amiga 3000/030 and the A500? 1. My 1989 Amiga 500 Rev6A comes with a Commodore 1084S monitor, the cost is about AUD $1400. In 1990, a 512 KB expansion card was fitted. 2. My dad purchased an ex-corporate A3000 (25Mhz Kickstart 2.04, 1MB Chip, 1 MB Fast) and sold the mentioned A500 in early 1992. The cost about $900. My family's annual computer hardware budget is around $1500 AUD. If this budget is not spent, it goes to a house mortgage.

Yes, we know that you were rich and your daddy gave you expensive toys: how many times do you have to repeat it?

And how many times I still have to tell you that this was NOT the case for the average Amigan?!?

Are you dumb?
Quote:

June 1992 in Australia [BIG IMAGE]

And you still fail to use images in the forum which don't break its layout, despite I've already told the same several times.

Are you aging badly? When will you learn how to correctly use the forum, when you're 6 feet below the graveyard?
Quote:

Quote: Do you understand that 1990 is the year were the ECS chipset was DELIVERED? So, they COMPLETED the development (which started BEFORE) and then sold the machines. Do you understand that 1989 era Amiga 500 Rev 6A has ECS Agnus? Except for 2MB Chip RAM addressing, 1989 era Amiga 500 Rev 6A's ECS Agnus 8372A has all the improvements like A3000's 8372AB. 1989 era Amiga 500 Rev 6 PCB design has 2MB Chip RAM jumpers and supports ECS Agnus 8372AB or 8372B. Product management reserved 2MB Chip RAM for the A3000.

What's not clear to you that when I write the ECS chipset I'm referring to ALL chips that are part of it and NOT only to Agnus?!?
Quote:

A3000's ECS Denise's four-color productivity modes are useless for games and inferior to VGA's 640x480p 16 colors.

And? Is it part of the ECS chipset or not? Because, do you know what: the article(s) is about the ECS Chipset: CHIP-SET.

Understood, or not?
Quote:

A3000's Amber with framebuffer is a workaround for flicker-free 640x400p 16 colors on Amiga chipset and PC VGA (31.5 kHz, 60 hz) monitor.

Who funny cares of it!!!

Do you understand that the topic was about the LOW-COST GA-MIN-G which was THE Amiga market?
Quote:

Quote: If something happened on the same year, then... it was already too late for the ECS release. Do you think that Commodore engineers had a time machine which allowed them to go back and fix / enhance ECS only because Wing Commander was released?!? Do you understand that the 1989 era Amiga 500 Rev 6 PCB design has 2MB Chip RAM jumpers? Do you understand that 1989 era Amiga 500 Rev 6A's ECS Agnus 8372A has all the improvements like A3000's 8372AB except 2MB Chip RAM addressing?

See above: CHIP-SET! CHIP-SET! CHIP-SET!
Quote:

Quote: "Oh, look: there's Wing Commander for PC now! Let's go back when we're working on the drawing board, then we can add the packed/chunky graphics and be ready for games like that. Amiga 2500UX's Amix has support for TIGA based A2410. The A2410 card is directly supported by the Amiga Unix SVR4 X11R4 from the X-Windows R4. Commodore's A2410 TIGA did nothing for Amiga chipset evolution.

LOL Yeah, it was very useful for the LOW-COST GAMING market, yes!!!


Quote:

Quote: How much they've costed compared to PCs and high-end machines like Mac-II (and successors) and X68000? This is the point. Do you understand it? Mastery of certain technology is equally important.

Sure. But more important is to do NOT go to bankrupt by losing your market of reference, isn't it?
Quote:

Quote: That's simply false: another invention. https://en.wikipedia.org/wiki/Amiga_Ranger_Chipset Jay Miner made sure that the Ranger chipset was completed and fully tested before he left Commodore in hope that one day the company would release it. But at that time, VRAM was considered expensive compared to DRAM, so Commodore refused to release Ranger for its high price which was unsuitable for the low-end systems like Amiga 500 to cover its costs. Dale Luck is in possession of a Ranger system prototype, but it's unknown what state it is in. https://web.archive.org/web/20090216124135/http://amigau.com/aig/comment-6.html Q: Will there ever be an advanced version of the chipset? A: Well, Jay Miner isn't working on anything right now... [RUMOR ALERT] The chip folks left in Los Gatos who are losing their lease in March were at one time thinking about 1k square 2meg chip space 128-color graphics, although still with 4 bit colour DACs though... and even stuff like a blitter per plane (!!) They were supposed to be done now, in the original plans; the chip designers will be gone in March, but the design may (?) continue in West Chester. Maybe they'll be here two years from now. Q: What will happen to the unused Los Gatos A2000 design? A: ?????? https://web.archive.org/web/20030506234259/http://amigau.com/aig/jayinterview2.jpg Jay Miner Interview Jay Miner: In spite of very limited manpower, we managed to finish the 1.2 software release; and design a revised set of custom chips for the next generation of Amiga computers .... They increase the display address range to two megabytes. These chips are completed and tested and only require a computer and memory to hold them together.

It was already discussed some times here: there's absolutely NO evidence of that! NOTHING! Not even a prototype.

The prototype was a fancy case which NOTHING else.

If it was only about attaching the memory (since the case was designed and available), why there's absolutely NOTHING which was shown working. Like... a boing ball, for example?
Quote:

1989 era Amiga 500 Rev 6A has 2 MB Chip RAM jumpers for 2MB ECS Agnus.

CHIP-SET! CHIP-SET! CHIP-SET!

How many times should I repeat to you, PARROT!
Quote:

Quote: It was expensive. DRAM's performance also improves over time.

CLAP CLAP CLAP!

What's not clear to you that I've already suggested to use the 14Mhz DRAMs which were available at the time?
Quote:

ET4000AX supports either 16-bit VRAM or 32-bit DRAM.

BOOOOM!

LOW-COST! LOW-COST! LOW-COST!
Quote:

SDRAM appeared in 1992 and SGRAM in 1994.

1990! 1990! 1990!
Quote:

Quote: Same with engineers, which had no vision of what was really needed for the PRIMARY market: games. The difference between 3DO (quadrilateral polygon 3D) vs Sony's Kutaragi-led PlayStation (triangle polygon 3D) teams. 3DO had a $699 price in 1993 which is similar to A500's $699 in 1987, but 3DO is not a general-purpose computer.

1990! 1990! 1990!

Let's see if I repeat it more times you'll finally get it into your brain (IF you've one, of course).
Quote:

Quote: That's garbage. The context was 1990! https://old.hotchips.org/wp-content/uploads/hc_archives/hc01/3_Tue/HC1.S8/HC1.8.2.pdf Sun GX Series Graphics Workstations in 1989. One of its key personnel co-founded NVIDIA along with an AMD microprocessor designer who is the current CEO. Curtis Priem co-founded NVIDIA with a history from IBM PGA (released in 1985, 4096 color palette, 256 colors) and Sun GX. He retired from NVIDIA in 2003. NVIDIA's NV1 was released in 1995. Market interest in the product quickly ended when Microsoft announced the DirectX specifications, based upon triangle polygon rendering. NVIDIA's intellectual foundation was from 1985 IBM PGA and late 1980s Sun GX. Sun's TEC features Common interface with SPARC, 680x0 and 80x86 51 MFLOPS single precision for transforms. Supports scaling, rotation, Integer, signed fixed point, binary, and IEEE single precision floating point.

MUHAHAHHAHAHA OMG: a Sun WORKSTATION now! A very low-cost GAMING machine!


Quote:

Meanwhile, Dave Haynie's AA3000+'s DSP3210 has 33 MFLOPS SP FP.

Oh, yeah. Why not? I bet it was costing only A LITTLE bit more the Amiga 500 of the time (1990), right?


Quote:

Quote: Even the PGA, but it was a BIT expensive to be used for games... Curtis Priem designed the 1985-era IBM PGA and it was expensive but it set the foundations for IBM's 256-color VGA in 1987 with many cost-reduced VGA cloners following this IBM standard.

PGA has NOTHING to do with the VGA!

Have you ever opened the hardware manual for both? It's clearly a funny NO in YOUR case!

Anyway, those were EXPENSIVE video cards. Understood, or not?
Quote:

Quote: They have both their responsibilities. Don't remove the engineers from the equation: they never understood what was really needed for the Amiga, as it can be clearly seen (same Blitter, same Paula). Sony's CEO listens to Kutaragi. Many of Kutaragi's meetings are with hardware prices and trade-offs.

What's not clear to you that my proposals were COST-EFFECTIVE and allowed to have a formidable low-cost machine which was super-competitive on the low-end videogames market and WAY better than AGA already on 1990?

This does NOT required people like Kutaragi, but the average engineers which were working for Commodore...

@cdimauro

Quote:

Which does NOT mean that we know what should have happened with the videogames in some years.

Gaming PC is one type of application on the PC. It's your mistake to assume that my argument is purely gaming PC.

The competitive pressure for faster VGA/SVGA products for productivity and Windows 2D acceleration also improves DOS gaming's 320x2x0 gaming resolution.

Intel has licensed NEC's µPD7220 as an 82720 graphics display controller, but it was crushed by IBM's VGA and 8514 standards which served as the basis for SVGA clones.


Quote:

But it wasn't aware of what could have happened in the future regarding videogames.

It's your mistake to assume my argument is purely gaming PC. Gaming PC is one type of application on the PC.

Again, the competitive pressure for faster VGA/SVGA models for productivity and Windows 2D acceleration also improves DOS gaming's 320x2x0 gaming resolution.

For the 1990s "32-bit 3D" generation, the RISC competition has scaled from game consoles to workstations.

1993-era Pentium's improved FPU benefits with games like Quake and productivity workloads. Quake almost single-handedly culled many X86 cloners from the X86 market.

The pattern is repeated with the gaming PC hardware industry is wreaking pure workstation vendors.

Quote:

Otherwise it wouldn't have spent BILLIONS on the Larrabee project, just to nominate another bug project which failed due to missing crystal balls...

https://youtu.be/rUODlNffWmU?t=1682 with the current head of Xbox Phil Spencer joining former Xbox head Peter Moore and Xbox creator Seamus Blackley on the subject of Intel's Paul Otellini intervention on using Intel CPUs instead of AMD on the original Xbox games console.

AMD and Microsoft completed the prototype Xbox before Intel's Paul Otellini intervention on using Intel CPUs.


https://venturebeat.com/games/making-of-the-xbox-1/view-all/

Similarly, Advanced Micro Devices thought it had a deal clinched to provide the microprocessor for the Xbox. That was true until the night before the announcement. Paul Otellini, one of the top executives at Intel, convinced Bill Gates that Intel could provide the chips. Intel clinched the deal and the AMD employees were dumbfounded when they heard the news at the announcement.


You have forgotten Intel's Paul Otellini's intervention with Bill Gates to use the Intel CPU ahead of the AMD CPU for the original Xbox.

The original Xbox project was before Intel's Larrabee project.

The Xbox team still has a corporate memory of backstabbing by Bill Gates that was pushed by Intel's Paul Otellini.

Quote:

It was already discussed some times here: there's absolutely NO evidence of that! NOTHING! Not even a prototype. The prototype was a fancy case which NOTHING else. If it was only about attaching the memory (since the case was designed and available), why there's absolutely NOTHING which was shown working. Like... a boing ball, for example?

There's a contradiction with Jay Miner's statement.

Jay Miner's 2 MB Chip RAM addressing feature is still on A500 Rev 6A's PCB design.

Regardless of the Amiga Ranger chipset's existence, there was a significant R&D stall after Amiga OCS.

Quote:

CHIP-SET! CHIP-SET! CHIP-SET! How many times should I repeat to you, PARROT!

Again, the 1989 A500 Rev6's ECS Agnus has all the features of A3000's ECS Agnus with only a single 1MB vs 2MB difference.

1989 A500 Rev6's 2 MB Chip RAM configuration is inherited in its design.

I don't base my arguments on retail release since A3000's 1990 ECS release could be delayed by other issues e.g. Super Buster issue.

"ECS Denise" resolution modes are useless for games and they are inferior to the competition.

Fact: ECS Agnus can operate without ECS Denise.

Quote:

CLAP CLAP CLAP! What's not clear to you that I've already suggested to use the 14Mhz DRAMs which were available at the time?

I already stated Commdore's 25 Mhz 32-bit memory controller mastery for A3000's 1990 release.

14 Mhz is a lower bar target. After the mastery, it's debate on costings.


https://archive.org/details/bub_gb_hqQJaNzN9IcC/page/n603/mode/2up
PC Mag 1992-08, page 604 of 664,
Diamond Speedstar 24 (ET4000AX ISA) has $169 USD retail. It's faster than a 14 Mhz video memory bus.

https://www.vgamuseum.info/index.php/cpu/item/552-diamond-speedstar-24-tseng-et4000ax
Diamond Speedstar 24 (ET4000AX ISA)
Memory clock speed: 40 Mhz
Core clock speed: 65 Mhz
Memory bus width: 32-bit
Memory type: Samsung's KM44C256CP-6 Fast Page DRAM with 60 ns access time.


Quote:

MUHAHAHHAHAHA OMG: a Sun WORKSTATION now! A very low-cost GAMING machine!

With existing 3D accelerator mastery from Sun GX, NVIDIA was able to design and release a 3D accelerator in around 2 years.

Nvidia was founded on April 5, 1993.
Nvidia released NV1 on May 22, 1995.

NVIDIA released NV1 in 1995 with the backing of Sega's Saturn ported games for the PC.
The PC market has strong competitors from 3DFX.

The big elephant in the room is Microsoft and its Direct3D specifications and GLQuake's OpenGL triangle 3D system killed NVIDIA's quadrilateral 3D NV1 and NV2.

RiVA128 (NV3) was released in 1997 and the turnaround was relatively quick from NV1 and NV2 mistakes.

Nvidia was founded on April 5, 1993.
Nvidia released NV1 on May 22, 1995. NVIDIA tried to fit Direct3D into NV1.
Nvidia released NV3 on August 25, 1997.

The difference is proven mastery over 3D accelerators and complex microprocessors from NVIDIA's co-founders.

----
Before Sega's Saturn project, Sega released Model 1 in 1992 and it's specifically designed for 3D polygon graphics. Model 1 arcade system board wasn't a low-cost GAMING machine, but it demonstrates Sega's certain mastery over 3D T&L hardware. Sega semicustomized Fujitsu TGP MB86233 DSP for 3D T&L hardware.

A 27-member group of Sega staff began development of the "Saturn" project.

By February 1994, talk of the Saturn was well underway, with developers supposedly receiving extremely early hardware reported to be vastly superior to the Amiga CD32, Atari Jaguar, and 3DO.

https://segaretro.org/History_of_the_Sega_Saturn/Development

Gaming PC needed to evolve into Pentium-level performance around 1995 and PC have Pentium since 1993 to establish a Pentium class install base.

Quote:

Oh, yeah. Why not? I bet it was costing only A LITTLE bit more the Amiga 500 of the time (1990), right?

I recall DaveH's statement that AT&T DSP3210 wasn't expensive.


Quote:

PGA has NOTHING to do with the VGA!

1. IBM PGC is partly register-compatible with CGA.
2. IBM PGC's 256-color (8-bit) design target is common with VGA.
3. IBM PGC's 640 × 480 target is common with VGA.

Hint: Use case.

Quote:

Have you ever opened the hardware manual for both? It's clearly a funny NO in YOUR case! Anyway, those were EXPENSIVE video cards. Understood, or not?

These specific use cases are repeated across IBM's products i.e. 8-bit colors, 640x480 resolution, CGA backward compatibility and' etc'.

The 2D acceleration use case is repeated for IBM 8514.

Without a specific use case design, you have Commodore's OCS/ECS is a good enough mindset.

The set of specific use cases should be strong enough to drive leadership's direction.

Last edited by Hammer on 14-Dec-2023 at 05:05 AM.
Last edited by Hammer on 14-Dec-2023 at 04:57 AM.
Last edited by Hammer on 14-Dec-2023 at 04:52 AM.
Last edited by Hammer on 14-Dec-2023 at 04:43 AM.

_________________
Ryzen 9 7900X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB
Amiga 1200 (Rev 1D1, KS 3.2, PiStorm32lite/RPi 4B 4GB/Emu68)
Amiga 500 (Rev 6A, KS 3.2, PiStorm/RPi 3a/Emu68)

@Hammer

Quote:

Hammer wrote: @cdimauro Quote: Which does NOT mean that we know what should have happened with the videogames in some years. Gaming PC is one type of application on the PC. It's your mistake to assume that my argument is purely gaming PC.

The CONTEXT is about THE Amiga market which was LOW-COST GAMING!
Quote:

The competitive pressure for faster VGA/SVGA products for productivity and Windows 2D acceleration also improves DOS gaming's 320x2x0 gaming resolution.

That's an absolute NON-SENSE! 2D acceleration have NOT improved DOS games: for some years it was only about software rendering, due to all differences of the graphics cards, even when the first VESA specs arrived.

DOOM arrived when? Does it uses any form for 2D hardware acceleration? Give an answer to those two simple questions and you'll how it was the REAL situation of the time!
Quote:

Intel has licensed NEC's µPD7220 as an 82720 graphics display controller, but it was crushed by IBM's VGA and 8514 standards which served as the basis for SVGA clones.

See above: IRRELEVANT!
Quote:

Quote: But it wasn't aware of what could have happened in the future regarding videogames. It's your mistake to assume my argument is purely gaming PC. Gaming PC is one type of application on the PC. Again, the competitive pressure for faster VGA/SVGA models for productivity and Windows 2D acceleration also improves DOS gaming's 320x2x0 gaming resolution.

PARROT MODE ON...
Quote:

For the 1990s "32-bit 3D" generation, the RISC competition has scaled from game consoles to workstations.

Ah, yes, it was clearly visible on the game systems of the time. Yes, yes...
Quote:

1993-era Pentium's improved FPU benefits with games like Quake and productivity workloads. Quake almost single-handedly culled many X86 cloners from the X86 market.

That was part of the CPU! The CONTEXT here was/is about the CHIPSET and, specifically, of the GRAPHIC/VIDEO part!

Even nowadays the CPU is having its big part on running videogames, but the GPU has the BIGGEST part.

How the Pentium-FPU has improved the VIDEO part of Quake will remain a mystery...

Last but not really least: 1990! 1990! 1990!
Quote:

The pattern is repeated with the gaming PC hardware industry is wreaking pure workstation vendors.

See above: IRRELEVANT.
Quote:

Quote: Otherwise it wouldn't have spent BILLIONS on the Larrabee project, just to nominate another bug project which failed due to missing crystal balls... https://youtu.be/rUODlNffWmU?t=1682 with the current head of Xbox Phil Spencer joining former Xbox head Peter Moore and Xbox creator Seamus Blackley on the subject of Intel's Paul Otellini intervention on using Intel CPUs instead of AMD on the original Xbox games console. AMD and Microsoft completed the prototype Xbox before Intel's Paul Otellini intervention on using Intel CPUs. https://venturebeat.com/games/making-of-the-xbox-1/view-all/ Similarly, Advanced Micro Devices thought it had a deal clinched to provide the microprocessor for the Xbox. That was true until the night before the announcement. Paul Otellini, one of the top executives at Intel, convinced Bill Gates that Intel could provide the chips. Intel clinched the deal and the AMD employees were dumbfounded when they heard the news at the announcement. You have forgotten Intel's Paul Otellini's intervention with Bill Gates to use the Intel CPU ahead of the AMD CPU for the original Xbox. The original Xbox project was before Intel's Larrabee project. The Xbox team still has a corporate memory of backstabbing by Bill Gates that was pushed by Intel's Paul Otellini.

First YOU reported about Andy (Grove): not me. It was Otellini? OK, np: doesn't change.

However the XBox team had NOTHING to do with the above. In fact, Intel provided ONLY the CPU. The chipset was provided by nVidia.

Intel had already experience with chipsets and, specifically, with video sections (and even video cards), but those were NOT the target.

In fact, Larrabee was started only SEVERAL YEARS AFTER the XBox project. It was a completely NEW decision about the company, which had nothing to with that (and BTW, I've worked with some colleagues which heavily contributed to this project. In fact, my primary work at Intel was exactly about the Xeon Phi family. The decision to jump on the high-end graphics video market was new, completely independent from the past, and had nothing to do with the consoles!).
Quote:

Quote: It was already discussed some times here: there's absolutely NO evidence of that! NOTHING! Not even a prototype. The prototype was a fancy case which NOTHING else. If it was only about attaching the memory (since the case was designed and available), why there's absolutely NOTHING which was shown working. Like... a boing ball, for example? There's a contradiction with Jay Miner's statement. Jay Miner's 2 MB Chip RAM addressing feature is still on A500 Rev 6A's PCB design.

What? Just using TWO MORE BITS for addressing more Chip Mem was... a revolution made by Miner? Is this the message that you want to give?!?
Quote:

Regardless of the Amiga Ranger chipset's existence, there was a significant R&D stall after Amiga OCS.

And?
Quote:

Quote: CHIP-SET! CHIP-SET! CHIP-SET! How many times should I repeat to you, PARROT! Again, the 1989 A500 Rev6's ECS Agnus has all the features of A3000's ECS Agnus with only a single 1MB vs 2MB difference. 1989 A500 Rev6's 2 MB Chip RAM configuration is inherited in its design.

See above: this was a VERY VERY VERY (three times to should be enough. I hope!) SMALL and OBVIOUS change to Agnus.
Quote:

I don't base my arguments on retail release since A3000's 1990 ECS release could be delayed by other issues e.g. Super Buster issue.

This has NOTHING to do with the chipset, which is completely independent from the CPU.
Quote:

"ECS Denise" resolution modes are useless for games and they are inferior to the competition.

That's why I've proposed something totally different, despite very easy to implement.
Quote:

Fact: ECS Agnus can operate without ECS Denise.

WoW! Mister Obvious!

But the CONTEXT was A BIT different, right? Do you recall it or not?
Quote:

Quote: CLAP CLAP CLAP! What's not clear to you that I've already suggested to use the 14Mhz DRAMs which were available at the time? I already stated Commdore's 25 Mhz 32-bit memory controller mastery for A3000's 1990 release. 14 Mhz is a lower bar target. After the mastery, it's debate on costings.

And guess what: the lower bar target was EXACTLY the Amiga market!

Which thanks to the 14Mhz DRAM could have had a BIG boost. Which is... THE CONTEXT!
Quote:

https://archive.org/details/bub_gb_hqQJaNzN9IcC/page/n603/mode/2up PC Mag 1992-08, page 604 of 664, Diamond Speedstar 24 (ET4000AX ISA) has $169 USD retail. It's faster than a 14 Mhz video memory bus. https://www.vgamuseum.info/index.php/cpu/item/552-diamond-speedstar-24-tseng-et4000ax Diamond Speedstar 24 (ET4000AX ISA) Memory clock speed: 40 Mhz Core clock speed: 65 Mhz Memory bus width: 32-bit Memory type: Samsung's KM44C256CP-6 Fast Page DRAM with 60 ns access time.

Hey, the above 14Mhz memory had 70ns access time. So, very very close to this 60 ns.

I assume then that there were several wait-states for accessing memory.

Whereas for the Amiga ECS++ I've planned zero delays (hence using 70ns memories).
Quote:

Quote: MUHAHAHHAHAHA OMG: a Sun WORKSTATION now! A very low-cost GAMING machine! With existing 3D accelerator mastery from Sun GX, NVIDIA was able to design and release a 3D accelerator in around 2 years. Nvidia was founded on April 5, 1993. Nvidia released NV1 on May 22, 1995. NVIDIA released NV1 in 1995 with the backing of Sega's Saturn ported games for the PC. The PC market has strong competitors from 3DFX. The big elephant in the room is Microsoft and its Direct3D specifications and GLQuake's OpenGL triangle 3D system killed NVIDIA's quadrilateral 3D NV1 and NV2. RiVA128 (NV3) was released in 1997 and the turnaround was relatively quick from NV1 and NV2 mistakes. Nvidia was founded on April 5, 1993. Nvidia released NV1 on May 22, 1995. NVIDIA tried to fit Direct3D into NV1. Nvidia released NV3 on August 25, 1997. The difference is proven mastery over 3D accelerators and complex microprocessors from NVIDIA's co-founders.

Again the crystal ball and now you move far away from THE CONTEXT: 1990 and LOW-COST GAMING...
Quote:

---- Before Sega's Saturn project, Sega released Model 1 in 1992 and it's specifically designed for 3D polygon graphics. Model 1 arcade system board wasn't a low-cost GAMING machine, but it demonstrates Sega's certain mastery over 3D T&L hardware. Sega semicustomized Fujitsu TGP MB86233 DSP for 3D T&L hardware. A 27-member group of Sega staff began development of the "Saturn" project. By February 1994, talk of the Saturn was well underway, with developers supposedly receiving extremely early hardware reported to be vastly superior to the Amiga CD32, Atari Jaguar, and 3DO. https://segaretro.org/History_of_the_Sega_Saturn/Development

SEGA Saturn was part of the NEW consoles generation. In fact, we're talking about 3-4 after the ECS release.

Could you tell which consoles had SEGA on 1990? And, in general, which consoles were available at the time?

Could you tell me their HARDWARE specs? CPUs, video processors, sound processors, memory used?

Because this is THE CONTEXT of the time. And Amiga had to fight against THAT market. Do you finally get it on your mind?!?
Quote:

Gaming PC needed to evolve into Pentium-level performance around 1995 and PC have Pentium since 1993 to establish a Pentium class install base.

This is ONLY THE CPU! There was and there is MUCH MORE besides the CPU, which is THE CONTEXT!!!
Quote:

Quote: Oh, yeah. Why not? I bet it was costing only A LITTLE bit more the Amiga 500 of the time (1990), right? I recall DaveH's statement that AT&T DSP3210 wasn't expensive.

Good. Then I assume that the Amiga 3000+ with this DSP could be sold at around the same prices of an Amiga 500 of the time (1990) to address the LOW-COST GAMING market which was THE marker fro the Amiga, right?

A plain YES or NO as answer is enough, thanks!
Quote:

Quote: PGA has NOTHING to do with the VGA! 1. IBM PGC is partly register-compatible with CGA. 2. IBM PGC's 256-color (8-bit) design target is common with VGA. 3. IBM PGC's 640 × 480 target is common with VGA. Hint: Use case.

Another complete non-sense: NO! Again, NO! The fact that the PGA had 256 colours and a 640x480 means NOTHING.

It was COMPLETELY DIFFERENT from the VGA!

And now tell me how to display 256 colours on 640x480 screen with the VGA. I'm super curious to see how you managed to do it (because you did it, right?).
Quote:

Quote: Have you ever opened the hardware manual for both? It's clearly a funny NO in YOUR case! Anyway, those were EXPENSIVE video cards. Understood, or not? These specific use cases are repeated across IBM's products i.e. 8-bit colors, 640x480 resolution, CGA backward compatibility and' etc'.

Which use case? The CONTEXT was the HARDWARE specs of PGA and VGA, which are completely different! They work differently!

Again, have you EVER opened their hardware manual and checked their registers and how they worked? It's clearly a big NO!

But you continue to talk about them, repeating the same things like a PARROT!
Quote:

The 2D acceleration use case is repeated for IBM 8514.

See above for this: it was USELESS for the games of the time (unless they used Windows. But DOS was the primary market for PC games, since several years).
Quote:

Without a specific use case design, you have Commodore's OCS/ECS is a good enough mindset.

LOL. Before that you said the ECS Denise was crap and now that the mindset was good.


Quote:

The set of specific use cases should be strong enough to drive leadership's direction.

Yeah, we saw it! THOUSANDS of games were released when the ECS chipset (CHIP-SET) appeared and thanks to that Commodore was clearly driving the leadership on the gaming market only because of that...

@cdimauro

Quote:

TWO years after the ECS and using... could you please tell me which kind of CPU, video and audio processors it was using on 1992? Just to see if you finally understand what means having a low-cost machine running games...

From October 1992, Australia's Your Computer magazine. Page 108 of 132, in AUD

TSENG Mega EVA/2 ( ET4000AX) = $230.
TSENG Mega EVA/1024 (ET4000AX) = $260 AUD. 1MB version.
These are TSENG Mega EVA with ISA ET4000 Sierra SC11486 HiColor DAC.

TSENG 24-bit MEGA 1024 = $300, 24-bit TrueColor DAC.

Trident SVGA 512 = $69.
Trident SVGA 1024 = $87

S3 Windows Accelerator = $400

PS; Diamond Speedstar 24 ET4000 board has a 24-bit DAC.

The ET4000AX-based SVGA solution was selected.

---------------

https://www.statista.com/statistics/1101900/unit-sales-snes-region/
North America: 22.88 million
Japan: 17.17 million
Europe: 8.15 million
Rest of the World: 0.9 million

Like Commodore's AGA platforms, SNES entered the European market in 1992 which is Commodore's core market.

Gaming PC has a new "32-bit" gaming experience not on SNES e.g. Q4 1993 Doom.

Nintendo workaround SNES's math power problem with built-in accelerator cards with games on a mass production scale e.g. low-cost RISC-based CPU SuperFX series.

Nintendo wasn't stupid like Commodore International's CEO.



Quote:

Hey, we know that you were rich and your daddy bought you several EXPENSIVE toys. But this was NOT the case for the average Amigan, which was able to afford the low-entry machines (yes: the ones which were most sold by Commodore). Have you got the concept or not? I don't know how else should I tell you

Australian Commodore and Amiga Review 1993-09, page 4 of 84, AUD

A1200 with 40MB HDD is $995.
GVP A1230 030/882 @ 40Mhz 4MB RAM = $1169

Total: $2,164

In 1992, A1200 was mostly missing in action in Australia.

For $2164 budget in 1993, it's a 486DX PC clone.

Page 28 of 84
$795 A1200 40 MB HDD with 1MB A500 trade in.

Total: $1,964 with GVP A1230 030/882 @ 40Mhz 4MB RAM.

----
Page 28 of 84
A4000/030 (120 MB HDD, 2MB Chip RAM) $1599 with A3000 (6MB RAM, 52 MB HDD) trade-in

486SX-25 or 486SX-33 PC clone price range in Xmas Q4 1993.
---


A500's larger market doesn't translate into A1200's small add-on market and A500 is missing AGA.

Why would an OCS/ECS Amiga 500 owner invest (e.g. GVP A530) in a dead-end gaming platform?

AGA should been released in 1990 or 1991 to build up a larger install base for A1200's CPU accelerator cards.

GVP is defunct in July 1995 since the A1200/A4000 addon market is tiny when compared to A500's install base.

http://www.bambi-amiga.co.uk/amigahistory/sales.html

----------

There are more than 4 million sold copies of PC Doom 1 and 2. IDsoftware estimated around 15 million copies floating i.e. piracy.

Gaming PCs are relatively large compared to the competition while it's a minority in the overall PC market.

Many of my primary school friends who owned Amiga 500s from 1988 to 1991 have changed platforms during the 1992 to 1993 time period.

Last edited by Hammer on 14-Dec-2023 at 06:52 AM.
Last edited by Hammer on 14-Dec-2023 at 06:45 AM.
Last edited by Hammer on 14-Dec-2023 at 06:44 AM.
Last edited by Hammer on 14-Dec-2023 at 06:39 AM.

_________________
Ryzen 9 7900X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB
Amiga 1200 (Rev 1D1, KS 3.2, PiStorm32lite/RPi 4B 4GB/Emu68)
Amiga 500 (Rev 6A, KS 3.2, PiStorm/RPi 3a/Emu68)

@cdimauro

Quote:

The CONTEXT is about THE Amiga market which was LOW-COST GAMING!

I have given you retail price quotes for the 1992 to 1993 time period for Australia and USA.


Quote:

That's an absolute NON-SENSE! 2D acceleration have NOT improved DOS games: for some years it was only about software rendering, due to all differences of the graphics cards, even when the first VESA specs arrived.

Your mistake is detaching fast 2D acceleration from fast frame buffer performance.

Fast 2D acceleration hardware is useless without a fast video frame buffer.

SVGA cards like ET4000 cover DOS gaming, DOS pro apps, and Windows 2D acceleration.

Other SVGA cards are also good choices such as Trident 8900CL, and WD90C32.

Quote:

DOOM arrived when?

Doom was released in December 1993 with magazine previews from around May 1993.

Doom was hyped by PC press and IDsoftware.

There are other PC games like Wing Commander 2 (1991), Star Wars X-Wing (March 1993), and IndyCar Racing (1993) besides Doom (1993).

IndyCar Racing level can be done on A1200 with 68030 @ 50Mhz.

Quote:

Does it uses any form for 2D hardware acceleration? Give an answer to those two simple questions and you'll how it was the REAL situation of the time!

Your mistake is detaching fast 2D acceleration from fast frame buffer performance when selected SVGA products support fast DOS gaming and Windows 2D acceleration.


Quote:

See above: IRRELEVANT!

RELEVANT.

Intel jumped on NEC's bandwagon and failed. IBM is the big elephant in the room to a certain point.

Quote:

PARROT MODE ON...

You can't handle the truth.

Quote:

First YOU reported about Andy (Grove): not me. It was Otellini? OK, np: doesn't change.

Intel management was pushing for the original Xbox game console contract after the prototype was completed.

Quote:

However the XBox team had NOTHING to do with the above. In fact, Intel provided ONLY the CPU.

Fact: Intel wanted to be involved with the Xbox game console project.

Xbox Team contracted AMD to provide the Duron CPU solution before Bill Gates's intervention that was pushed by Intel management.

AMD and NVIDIA created the original Xbox prototype before Intel's Pentium III-128 L2 cache.

The pain is still in the memory of the Xbox team.

Quote:

The chipset was provided by nVidia.

NVIDIA wasn't the problem during the prototype phase.

Quote:

That was part of the CPU! The CONTEXT here was/is about the CHIPSET and, specifically, of the GRAPHIC/VIDEO part!

Wrong. The pain is still in the memory of the Xbox team.


Quote:

Even nowadays the CPU is having its big part on running videogames, but the GPU has the BIGGEST part.

Both components are important since the CPU is the director and maintains the BVH structure.

GPU's hardware RT is fed with BVH structure segments from the CPU. CPU maintains BVH geometry structure in relation to game simulation.

A high-performance CPU is important for maintaining BVH structure and related game logic simulation. CyberPunk 2077 DXR hammers both CPU and GPU.

GPU is a slave that paints the CPU's viewport.


Quote:

How the Pentium-FPU has improved the VIDEO part of Quake will remain a mystery...

3D render has a pipeline. Fast math power needs a fast-frame buffer.

Quote:

Last but not really least: 1990! 1990! 1990!

The situation in 1993 had its origins in 1990 or before 1990 e.g. ET4000AX and 486 were released in 1989.

Motorola and Apple delivered cost-performance competitive 68LC040 CPU in Q4 1993.

Quote:

But the CONTEXT was A BIT different, right? Do you recall it or not?

Regardless, 128 colors with 7-bit planes wouldn't be enough.

Quote:

Intel had already experience with chipsets and, specifically, with video sections (and even video cards), but those were NOT the target.

Gaming PC is just one of many application types and use cases for PCs.


Quote:

In fact, Larrabee was started only SEVERAL YEARS AFTER the XBox project. It was a completely NEW decision about the company, which had nothing to with that (and BTW, I've worked with some colleagues which heavily contributed to this project. In fact, my primary work at Intel was exactly about the Xeon Phi family. The decision to jump on the high-end graphics video market was new, completely independent from the past, and had nothing to do with the consoles!). [quote] Intel's Larrabee GPU project has failed and it was replaced by Intel's ARC. Intel 144-core Sierra Forest effectively continues the many-core Larrabee CPU project as Intel's many-core E-Core alternative to AMD's 128-core/256 threads Bergamo. [quote] Good. Then I assume that the Amiga 3000+ with this DSP could be sold at around the same prices of an Amiga 500 of the time (1990) to address the LOW-COST GAMING market which was THE marker fro the Amiga, right? A plain YES or NO as answer is enough, thanks! [quote] FYI, AA3000+ has the 68040 killer DSP solution on 25 Mhz 32-bit Fast RAM. Placing DSP3210 on slow Chip RAM is LOL. What's your intended memory configuration? [quote] SEGA Saturn was part of the NEW consoles generation. In fact, we're talking about 3-4 after the ECS release. Could you tell which consoles had SEGA on 1990? And, in general, which consoles were available at the time? Could you tell me their HARDWARE specs? CPUs, video processors, sound processors, memory used? Because this is THE CONTEXT of the time. And Amiga had to fight against THAT market. Do you finally get it on your mind?!?

Sega's Model 1 R&D was completed before the product release and games were also created.

Sega's solution was five DSPs @ 16 Mhz and custom raster graphics that exceeded Mega Drive's.

Sega purchased off-the-shelf 68000 and DSPs for Model 1 and SuperH for Saturn. The important part is the games. Sega evaluated 68030 CPU for the Saturn and they didn't want $500 console.

Last edited by Hammer on 14-Dec-2023 at 08:16 AM.

_________________
Ryzen 9 7900X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB
Amiga 1200 (Rev 1D1, KS 3.2, PiStorm32lite/RPi 4B 4GB/Emu68)
Amiga 500 (Rev 6A, KS 3.2, PiStorm/RPi 3a/Emu68)

@Hammer

Quote:

Hammer wrote: @cdimauro Quote: TWO years after the ECS and using... could you please tell me which kind of CPU, video and audio processors it was using on 1992? Just to see if you finally understand what means having a low-cost machine running games... From October 1992, Australia's Your Computer magazine. Page 108 of 132, in AUD TSENG Mega EVA/2 ( ET4000AX) = $230. TSENG Mega EVA/1024 (ET4000AX) = $260 AUD. 1MB version. These are TSENG Mega EVA with ISA ET4000 Sierra SC11486 HiColor DAC. TSENG 24-bit MEGA 1024 = $300, 24-bit TrueColor DAC. Trident SVGA 512 = $69. Trident SVGA 1024 = $87 S3 Windows Accelerator = $400 PS; Diamond Speedstar 24 ET4000 board has a 24-bit DAC. The ET4000AX-based SVGA solution was selected.

That's only the video card. What about complete configuration of a gaming PC AND the price of an Amiga 500?
Quote:

https://www.statista.com/statistics/1101900/unit-sales-snes-region/ North America: 22.88 million Japan: 17.17 million Europe: 8.15 million Rest of the World: 0.9 million Like Commodore's AGA platforms, SNES entered the European market in 1992 which is Commodore's core market.

SNES was sold in 1990, not 1992. So, same year of the ECS.

Could you list its hardware specs?
Quote:

Gaming PC has a new "32-bit" gaming experience not on SNES e.g. Q4 1993 Doom.

Again: the context was 1990! 1990! 1990! The same year that the SNES was released...
Quote:

Nintendo workaround SNES's math power problem with built-in accelerator cards with games on a mass production scale e.g. low-cost RISC-based CPU SuperFX series. Nintendo wasn't stupid like Commodore International's CEO.

Non-sense: that's an ADDITIONAL cartridge (a new chip in the cartridge, to be more precise) which is the equivalent of an accelerator card for Amigas.
Quote:

Quote: Hey, we know that you were rich and your daddy bought you several EXPENSIVE toys. But this was NOT the case for the average Amigan, which was able to afford the low-entry machines (yes: the ones which were most sold by Commodore). Have you got the concept or not? I don't know how else should I tell you Australian Commodore and Amiga Review 1993-09, page 4 of 84, AUD A1200 with 40MB HDD is $995. GVP A1230 030/882 @ 40Mhz 4MB RAM = $1169 Total: $2,164 In 1992, A1200 was mostly missing in action in Australia. For $2164 budget in 1993, it's a 486DX PC clone. Page 28 of 84 $795 A1200 40 MB HDD with 1MB A500 trade in. Total: $1,964 with GVP A1230 030/882 @ 40Mhz 4MB RAM. ---- Page 28 of 84 A4000/030 (120 MB HDD, 2MB Chip RAM) $1599 with A3000 (6MB RAM, 52 MB HDD) trade-in 486SX-25 or 486SX-33 PC clone price range in Xmas Q4 1993. ---

Again, useless PADDING, since the CONTEXT was 1990!
Quote:

A500's larger market doesn't translate into A1200's small add-on market and A500 is missing AGA. Why would an OCS/ECS Amiga 500 owner invest (e.g. GVP A530) in a dead-end gaming platform?

In fact it wouldn't have needed having an Amiga 500 with ECS++ on 1990. As per the CONTEXT of the discussions (and the articles).

Got the point after dozen and dozen of times that I've written it, or it still not enough for your brain?
Quote:

AGA should been released in 1990 or 1991 to build up a larger install base for A1200's CPU accelerator cards.

AGA was too little even if it would have been released on 1990, as I've clearly proven.
Quote:

GVP is defunct in July 1995 since the A1200/A4000 addon market is tiny when compared to A500's install base. http://www.bambi-amiga.co.uk/amigahistory/sales.html

Amen. But... WHO CARES? The CONTEXT was Low-End GAMING market on 1990.
Quote:

There are more than 4 million sold copies of PC Doom 1 and 2. IDsoftware estimated around 15 million copies floating i.e. piracy. Gaming PCs are relatively large compared to the competition while it's a minority in the overall PC market.

Maybe because they were expensive for the low-end gaming market? Just an idea, eh!
Quote:

Many of my primary school friends who owned Amiga 500s from 1988 to 1991 have changed platforms during the 1992 to 1993 time period.

Thanks for having shared an useless part of your story...
Quote:

Hammer wrote: @cdimauro Quote: The CONTEXT is about THE Amiga market which was LOW-COST GAMING! I have given you retail price quotes for the 1992 to 1993 time period for Australia and USA.

USELESS, since the context is 1990. 1990! 1990!
Quote:

Quote: That's an absolute NON-SENSE! 2D acceleration have NOT improved DOS games: for some years it was only about software rendering, due to all differences of the graphics cards, even when the first VESA specs arrived. Your mistake is detaching fast 2D acceleration from fast frame buffer performance.

Maybe because they are completely different things and you can have just of them or even both?
Quote:

Fast 2D acceleration hardware is useless without a fast video frame buffer.

LOL! Then kick out the Blitter from all Amigas, because it's useless without a fast frame buffer.
Quote:

SVGA cards like ET4000 cover DOS gaming, DOS pro apps, and Windows 2D acceleration. Other SVGA cards are also good choices such as Trident 8900CL, and WD90C32.

Usual PADDING...
Quote:

Quote: DOOM arrived when? Doom was released in December 1993 with magazine previews from around May 1993. Doom was hyped by PC press and IDsoftware. There are other PC games like Wing Commander 2 (1991), Star Wars X-Wing (March 1993), and IndyCar Racing (1993) besides Doom (1993). IndyCar Racing level can be done on A1200 with 68030 @ 50Mhz.

And all of them used 2D Acceleration, right? As per your previous statement, they should...
Quote:

Quote: Does it uses any form for 2D hardware acceleration? Give an answer to those two simple questions and you'll how it was the REAL situation of the time! Your mistake is detaching fast 2D acceleration from fast frame buffer performance when selected SVGA products support fast DOS gaming and Windows 2D acceleration.

See above: irrelevant. A coprocessor can be very useful EVEN without a fast frame buffer.

This again clarified, which DOS games have used the 2D accelerators built in the SVGA cards?

You mentioned DOOM before, when talking about it: have it used some 2D accelerator? Yes or no, please.
Quote:

Quote: See above: IRRELEVANT! RELEVANT. Intel jumped on NEC's bandwagon and failed. IBM is the big elephant in the room to a certain point.

And both remain IRRELEVANT considered the context...
Quote:

Quote: PARROT MODE ON... You can't handle the truth.

No, I can. I "just" can't handle the bullshits...
Quote:

Quote: First YOU reported about Andy (Grove): not me. It was Otellini? OK, np: doesn't change. Intel management was pushing for the original Xbox game console contract after the prototype was completed.

And what's YOUR problem? It was interested on selling its CPUs (C-P-U)!
Quote:

Quote: However the XBox team had NOTHING to do with the above. In fact, Intel provided ONLY the CPU. Fact: Intel wanted to be involved with the Xbox game console project. Xbox Team contracted AMD to provide the Duron CPU solution before Bill Gates's intervention that was pushed by Intel management. AMD and NVIDIA created the original Xbox prototype before Intel's Pentium III-128 L2 cache.

Exactly, and? It was only the CPU!
Quote:

The pain is still in the memory of the Xbox team.

No, it's in your a$$ only because you're an AMD die hard blind fanatical, so you're still burning for this.
Quote:

Quote: The chipset was provided by nVidia. NVIDIA wasn't the problem during the prototype phase.

Neither the CPU, which was a regular P3 one (only with half the L2 cache).
Quote:

Quote: That was part of the CPU! The CONTEXT here was/is about the CHIPSET and, specifically, of the GRAPHIC/VIDEO part! Wrong. The pain is still in the memory of the Xbox team.

No, it's YOUR problem because you're still burning for it.

This clarified, again: the context was the CHIP-SET! NOT the CPU!
Quote:

Quote: Even nowadays the CPU is having its big part on running videogames, but the GPU has the BIGGEST part. Both components are important since the CPU is the director and maintains the BVH structure. GPU's hardware RT is fed with BVH structure segments from the CPU. CPU maintains BVH geometry structure in relation to game simulation. A high-performance CPU is important for maintaining BVH structure and related game logic simulation. CyberPunk 2077 DXR hammers both CPU and GPU. GPU is a slave that paints the CPU's viewport.

It remains the most important component compared to the CPU. It's so much important that even an AMD die hard blind fanatical like you is using nVidia GPUs instead the ones of your beloved AMD...
Quote:

Quote: How the Pentium-FPU has improved the VIDEO part of Quake will remain a mystery... 3D render has a pipeline. Fast math power needs a fast-frame buffer.

LOL: TOTAL NON-SENSE! The FPU does NOT use the graphic card's fast-frame buffer for its calculations! There's NOTHING that could help the CPU's FPU for processing its 3D calculations.

You don't know of what you talk about!!!
Quote:

Quote: Last but not really least: 1990! 1990! 1990! The situation in 1993 had its origins in 1990 or before 1990 e.g. ET4000AX and 486 were released in 1989. Motorola and Apple delivered cost-performance competitive 68LC040 CPU in Q4 1993.

And here comes back the famous Hammer's crystal ball...

Another total non-sense, of course, since chips sold on 1990 were designed years before, and of course targeting 1990 and NOT 1993 or something like that.

You've absolutely no clue of how chips are designed and go to the market!
Quote:

Quote: But the CONTEXT was A BIT different, right? Do you recall it or not? Regardless, 128 colors with 7-bit planes wouldn't be enough.

CLAP CLAP CLAP.

Where are the Ranger's demos? Wasn't it enough to attach the memory to the chipset, since it was "completed"? Show me it at work!
Quote:

Quote: Intel had already experience with chipsets and, specifically, with video sections (and even video cards), but those were NOT the target. Gaming PC is just one of many application types and use cases for PCs.

Gaming PC wasn't the primary interest of Intel for its graphics chipset. Specifically, it wasn't at all on the CONTEXT (AKA 1990).
Quote:

Quote: In fact, Larrabee was started only SEVERAL YEARS AFTER the XBox project. It was a completely NEW decision about the company, which had nothing to with that (and BTW, I've worked with some colleagues which heavily contributed to this project. In fact, my primary work at Intel was exactly about the Xeon Phi family. The decision to jump on the high-end graphics video market was new, completely independent from the past, and had nothing to do with the consoles!). Intel's Larrabee GPU project has failed and it was replaced by Intel's ARC.

Again, you've no idea of what you talk about. Larrabee was a failure, yes, and it was replaced by the Xeon Phi: NOT Arc!

Only in the recent years Arc has replaced Xeon Phi.
Quote:

Intel 144-core Sierra Forest effectively continues the many-core Larrabee CPU project as Intel's many-core E-Core alternative to AMD's 128-core/256 threads Bergamo.

Those are completely different processors which have NOTHING to do with Larrabee, besides having a lot of cores!

You've no idea of why Larrabee was designed as it was, and of course you know absolutely nothing of its internal architecture (which wasn't even x64-compatible!)
Quote:

Quote: Good. Then I assume that the Amiga 3000+ with this DSP could be sold at around the same prices of an Amiga 500 of the time (1990) to address the LOW-COST GAMING market which was THE marker fro the Amiga, right? A plain YES or NO as answer is enough, thanks! FYI, AA3000+ has the 68040 killer DSP solution on 25 Mhz 32-bit Fast RAM. Placing DSP3210 on slow Chip RAM is LOL. What's your intended memory configuration?

1MB. Now tell me how much all of this would have costed compared to an A500...
Quote:

Quote: SEGA Saturn was part of the NEW consoles generation. In fact, we're talking about 3-4 after the ECS release. Could you tell which consoles had SEGA on 1990? And, in general, which consoles were available at the time? Could you tell me their HARDWARE specs? CPUs, video processors, sound processors, memory used? Because this is THE CONTEXT of the time. And Amiga had to fight against THAT market. Do you finally get it on your mind?!? Sega's Model 1 R&D was completed before the product release and games were also created. Sega's solution was five DSPs @ 16 Mhz and custom raster graphics that exceeded Mega Drive's. Sega purchased off-the-shelf 68000 and DSPs for Model 1 and SuperH for Saturn. The important part is the games. Sega evaluated 68030 CPU for the Saturn and they didn't want $500 console.

Do you understand that with this you're already answered yourself? Those gaming systems should be LOW-COST!

So, NOT the target for the new Amiga which should have replaced the A500 for this market (which was THE market of reference).

@Hammer

Quote:

Hammer wrote: TSENG Mega EVA/1024 (ET4000AX) = $260 AUD. 1MB version. Trident SVGA 512 = $69.

This is why so many PC clones came with a crappy trident VGA card.

Quote:

June 1992 in Australia

Oh yeah a 386DX-40 for only $1320 - with that crappy 512k VGA card whose brand name we won't mention, and a miserable 1MB of RAM. No sound card. No hard drive. No operating system. This is only half a PC! Add all the stuff you need to make it usable and now what does it cost?

But of course the real killer is - it doesn't run Amiga software! Doesn't connect to a TV either - so much for playing games on the big screen in the living room.

Quote:

A1200 with 40MB HDD is $995.

And that's all I need for now. Maybe buy a RAM board or accelerator card later when prices come down and games start needing it. But will I still be using it 30 years later? Will it hold its value? Hell yeah!

But that 386DX will be worth even more, right? What you say, 2 years later the CMOS battery leaked over the motherboard and killed it? And I had to buy a new PC anyway because even with 8MB it was too slow running Windows 95? Besides which PC games now need at least a 486 or preferably a Pentium? (can't just upgrade the CPU and motherboard because the case and power supply are too small). And then 2 years later I will have to do it all again? Old PCs are piling up now, another trip to the dump may be in order...

@bhabbott

In 1993, my Amiga computing friend, sold his expanded A500 setup and bought a 486 MS-DOS PC.
Powerful? Yes.
Expensive? Even as a clone, yes.
Expandable? Hell yes.
Ugly? You bet.
Monotasking? Yep.
Sound config issue with games? Count on it.
Interrupt issues when adding new hardware? 10-4.
Lock up the system by holding down a mouse button? True.
Decoding eight character file names? ROGER.
Constantly re-calibrating the analog joystick? Check.

I once brought over my A1200 and he put its multi-tasking to the test.
Downloading a DMS file from a BBS, while simultaneously writing another DMS to a Floppy, while playing a MOD file, while pulling down a screen to continue creating in Deluxe Paint.

This was just the stock EC020 with a RAM expansion including 4MB RAM + 68882.
The 486 would go on to run Games like Doom and Doom II. The A1200 could not match it in raw power, but it outmatched it in user experience, flexibility, and multimedia creativity.

In the A500 days, I made fun of PC green-screen blip-bloop gaming, and I would recommend the A500 to anyone who cared to listen. But I never recommended the A1200 to anyone looking to buy their first computer. To me it made sense as an upgrade path from a previous Amiga. AGA was a big step up from OCS, and by the looks of it even ECS. It provided me a solid 5 years of visual fun and creativity until I got the Permedia 2 Bvission RTG addon card.

_________________
All the way, with 68k

@Hammer

this video is really interesting.
few second more it show how much time was wasted on c2p.
if aga has chunky pixels doom wil run at least two times faster.
and be playable and commodore will not bankrupt in 1994

commodore bankrupt because aga has not chunky pixels.
nobody know if commodore bankrupt if they have two years more and finish hombre.
speculations that commodore have to bankrupt is pure bs based on nothing.