@NutsAboutAmiga

LOL! His "bit" is definitly most significant!

_________________
Doing stupid things for fun...

@Hammer

Quote:

IBM still has a software consultancy business in the billions of dollars.

Right, and what’s that got to do with them not selling PC hardware? I merely pointed out that it wasn’t just the “consumer� PC division that went to Lenovo, but also their server PC devision. The computer hardware they sell now are all based around z and POWER. What this has to do with their consulting activities and phony practices in Australia… I really don’t know. IBM also owns RedHat (or vece versa), at least that’s a little x86 relevant.

_________________
B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC

@Hammer

Quote:

Hammer wrote: ARM is a direct response to Commodore's crap 65xx CPU roadmap.

If it wasn't for Commodore buying MOS the 6502 probably would have died in 1976.

With no 6502, what would Acorn put in the BBC Micro? Probably a Z80, since the BBC had CP/M as one of its requirements. Then it wouldn't have needed The Tube to connect a second processor, and they probably wouldn't have developed the ARM CPU for it.

So you are mostly right. since without Commodore ARM probably wouldn't exist.

Except for the 'crap roadmap' part.

Commodore's 'roadmap' for the 6502 was to use it in their machines, with variations to suit - eg. the 6509 (used in the CBM-II) which ran at 2MHz and could address 1MB. Other manufacturers produced their own variants. In 1981 the Western Design Center created the 65C02, then in 1985 the 16-bit 65C816 (used in the Apple IIGS).

Realistically that's about as far as it was worth pushing the 6502. There was no demand for a higher performance 6502 based CPU for desktop computers because that role was taken by x86 in the PC, a result of IBM choosing it in 1981. That only left the embedded market where compatibility wasn't important - opening the door to an entirely new architecture. Acorn's ARM eventually did well there due to its high electrical efficiency in applications that needed the compute power of a 32-bit CPU, starting with smartphones in the early 2000s.

@kolla

"My A3000 has a CSPPC/CVPPC.

The CPU is on the CSPPC
The RAM is on the CSPPC
The video output is on the CVPPC
The storage is on the CSPPC UW-SCSI

So what's used on the A3000 motherboard?
Well, the keyboard and mouse attach there.

And I have a DENEB Z3 card for, for USB and networking (to a pi heh), but... if there was a "pro" version of BlueSCSI with UW-SCSI capabilities/speeds, I'd happily move networking to that instead, and just remove the Deneb.

(btw, this A3000 has OS4FE and old MorpOS installed, but defaults to OS 3.2.2.1 with WarpUP)"

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


That is too funny, I remember in the 90's a group of us were laughing at just that.

All the almight 3000D was doing was minimal, with RTG, The 3040 handling SCSI, Memory, and CPU. The mightily 3000 was just to slow and you had to look to the add ons to avoid it.

All the Amiga became was just another PC, albeit slower. LOL. Still loved and used regardless even today, even though the awesome chipset is largely ignored...

Last edited by Matt3k on 20-Jan-2025 at 06:48 PM.

@Matt3k

Quote:

even though the awesome chipset is largely ignored...

Well, compared to what?

CIAA and CIAB chips are not complicated state of art chips, they first introduced in the C64 maybe before that. Its 8bit chip is mounted on a 16bit bus, and later to 32bit data bus.

Etch CIA chip are simple digital counters, that can be connected in serial.

Paula chip is not complicated, what makes it good,
it is good audio separation, and low noise.

The AGA/OCS/ECS graphic card is outdated.

But chips are necessary, because the software does need this stuff. Amiga software is tied to the hardware, like it or not… but that’s how it is.

just for programs, CIAA/CIAB and Paula chip is necessary, for games you just can’t play games without the complete AGA chipset (Alice).

I believe Draco did have CIAA/CIAB/Paula, but lacked some other chips.

Last edited by NutsAboutAmiga on 26-Jan-2025 at 01:37 PM.
Last edited by NutsAboutAmiga on 20-Jan-2025 at 07:23 PM.
Last edited by NutsAboutAmiga on 20-Jan-2025 at 07:22 PM.

_________________
http://lifeofliveforit.blogspot.no/
Facebook::LiveForIt Software for AmigaOS

@kolla

Quote:

kolla wrote: @Hammer Quote: IBM still has a software consultancy business in the billions of dollars. Right, and what’s that got to do with them not selling PC hardware? I merely pointed out that it wasn’t just the “consumer� PC division that went to Lenovo, but also their server PC devision. The computer hardware they sell now are all based around z and POWER. What this has to do with their consulting activities and phony practices in Australia… I really don’t know. IBM also owns RedHat (or vece versa), at least that’s a little x86 relevant.

IBM's software consulting service can lead to hardware sales. Hardware without business process software is useless for business.

_________________
Amiga 1200 (rev 1D1, KS 3.2, PiStorm32/RPi CM4/Emu68)
Amiga 500 (rev 6A, ECS, KS 3.2, PiStorm/RPi 4B/Emu68)
Ryzen 9 7950X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB

@bhabbott

Quote:

If it wasn't for Commodore buying MOS the 6502 probably would have died in 1976. With no 6502, what would Acorn put in the BBC Micro? Probably a Z80, since the BBC had CP/M as one of its requirements. Then it wouldn't have needed The Tube to connect a second processor, and they probably wouldn't have developed the ARM CPU for it.

Since Jack Tramiel is marketing 65xx CPUs like soda drinks, MOS's 65xx key engineers have exited CGS e.g.
1. Bill Mensch formed the Western Design Center.
2. Rod Orgill joined Hewlett Packard.

Before Acorn's BBC Micro, Acorn Atom used MOS 6502 and this microcomputer was released in 1980. https://en.wikipedia.org/wiki/Acorn_Atom

Read https://en.wikipedia.org/wiki/Acorn_System for early Acorn systems before Acorn's 1980 Atom release and they used MOS 6502.

https://www.commodore.ca/commodore-history/the-rise-of-mos-technology-the-6502/
According to Bill Mensch, MOS/CSG had a working 12 Mhz 6502 prototype before Motorola released 12 Mhz 68000 in 1983. Jack Tramiel rejected the 12 Mhz 6502 evolution.

The founding of MOS was due to Motorola's corporate behavior.

The founding of AMD and Intel were due to Motorola's corporate behavior being imported into Fairchild Semiconductor.

Around the end of 1974 to 1975, one of the key designers for 6800, Bill Lattin joined Andy Grove's Intel (ref 1).

The root cause is Motorola, followed by Jack Tramiel.

Reference
1. https://thechipletter.substack.com/p/leaving-arizona
This web link has a story about Motorola's corporate behavior during 1974 and 1975.
The customer (system integrator)'s biggest concern wasn’t the 6800’s architecture, but its price, closely matching the price of the 8-bit Intel's 8080.

Motorola had a behavior using Intel as #metoo price guide e.g. Motorola's full 32bit 68K following Intel's 386 prices until it was priced out by AMD's Am386-40 in 1991-1992.

Around 1986, SGI (system integrator) made similar remarks about Motorola's performance vs price issues with 68K.


Quote:

Realistically that's about as far as it was worth pushing the 6502. There was no demand for a higher performance 6502 based CPU for desktop computers because that role was taken by x86 in the PC, a result of IBM choosing it in 1981.

Like IBM DisplayWrite, IBM's TopView has failed.

SuperPET's 16-bit ALU math equipped 6809 debunks your argument since Jack Tramiel rejected 16-bit ALU equipped 65xx R&D e.g. the main reason for Bill Mensch exiting MOS/CSG and formed the Western Design Center.

In terms of unit sales up to 1985, Commodore's C64 was keeping up with IBM PC.

Jack Tramiel didn't leverage C64's economies of scale with incompatible SuperPET. This Tramiel product segmentation mindset would be applied to C16/Plus4 vs C64, Atari ST vs MegaST (with blitter in late 1986). With collapsing sales, Tramiel's Atari STE (with blitter) response in 1989 was too late.

C128 wasn't the new standard C64 since C128's C900 graphics chipset (MOS 8563) wasn't integrated with the C64's VIC-II for cost reduction.

Due to Jack Tramiel's actions, there was a brain drain after C64's release.

Unlike the C64, the IBM PC's MDA high-resolution text mode is suitable for killer DOS apps like Lotus 123 and Word Perfect. VisiCalc died along with platforms such as DRi CP/M, CBM (8bit business microcomputers), and 'etc'. Lotus 123 2.0 supported X87 FPUs.

CSG/MOS 65xx CPU is a dead end, hence Acorn's 32-bit ARM existence.

Jack Tramiel's SuperPET's 6809 (with 16-bit ALU math) token selection shows the market that the CSG/MOS 65xx CPU is a dead end.

C128's Z80 CP/M support in 1985 was overshadowed by DRi's GEM (graphics GUI)/GEMDOS on Atari ST, Apple Mac 512K (GUI), Microsoft Windows 1.0 (GUI) and Commodore's Amiga 1000 (GUI). 1985 for Apple Mac 512K was about next-gen GUI killer app releases.

During 1984-1985, ARM, 80386, and 68020 were 32-bit CPU releases while Western Design Center released 16-bit ALU-equipped WDC65816.

Intel wisely ganged up with Compaq, Microsoft, and SCO to quickly execute 386's system integration R&D for the 1986 (Compaq's AT386 standard) and 1987 (Windows 2.x 386, Xenix 386) releases.

#Metoo R&D = behind the competitors.


-------------
CSG's C65 group recycled Amiga's bitplane graphics IP to create C65's 8-bitplane (256 colors) before Herni Rubin's slow-paced R&D Amiga group. One of the key engineers for C65's VIC-III quickly designed the AA Lisa 8 bitplane (256 colors), AAA's 16.7 million color palette, and 64-bit fetch IP (two 32bit double pump) for the ECS Agnus variant (Alice).


Quote:

That only left the embedded market where compatibility wasn't important - opening the door to an entirely new architecture. Acorn's ARM eventually did well there due to its high electrical efficiency in applications that needed the compute power of a 32-bit CPU, starting with smartphones in the early 2000s.

ARM pushed out Motorola's 68000-based Dragon Ball from the handheld market.

In the 2000s, ARM, MIPS, SuperH, and Intel Atom competed in the handheld market.

Quote:

Except for the 'crap roadmap' part.

You can't handle the truth.

Quote:

Commodore's 'roadmap' for the 6502 was to use it in their machines, with variations to suit - eg. the 6509 (used in the CBM-II) which ran at 2MHz and could address 1MB. Other manufacturers produced their own variants. In 1981 the Western Design Center created the 65C02, then in 1985 the 16-bit 65C816 (used in the Apple IIGS).

Apple IIGS was released on the 15th of September 1986. You're wrong.
https://en.wikipedia.org/wiki/Apple_IIGS

65816 wasn't a Commodore product.

CBM-II has an 8-bit ALU-equipped CPU (i.e. CGS 6509, optional Z80/8080), C64's VIC-II, and Motorola 6845 display controller (clone of Hitachi HD46505).

Meanwhile, IBM's original PC has optional 8087 FPU for native processing for INT32, INT64, FP32, FP64 and FP80 datatypes. Intel 8088 has 16-bit ALU.

16bit and 32bit integers are emulated on 6502 CPUs, hence the reason for SuperPET's 6809 selection with 16bit ALU math.

X87 sets the pattern for follow-on X87 FPUs that can be used for longer datatype integer processing which is important in Pentium-era games e.g. PC's Tomb Raider DOS version.

Against Intel 8086, Motorola has 6809 which includes 16-bit ALU math. Intel released 8088 in 1979.

Motorola wouldn't counter Intel's X87 FPUs (8087, 80287) until 1984's 68881 release.

SUN-3's 68020/68881/proprietary Sun MMU combo is not a mass-produced box due to its very expensive price. Apple Mac II's 68020/68881 combo was released in 1987.

Intel 386 has a standard MMU and every 386 CPU includes an integrated MMU i.e. Intel went beyond #metoo.

Western Design Center's 16bit ALU 65816 was a #metoo R&D against Intel's 16bit ALU equipped 8086 (released in 1978) and 80286 (released in 1982).

Western Design Center's 65816 design parameters were frozen in time during Intel's 16-bit CPU release window.

For 32-bit ARM R&D, Acorn skips Intel's 16-bit 80286 CPU release window. VTI (VLSI Tech Inc) supported ARM R&D.

Western Design Center's 65816 has mass production issues with C= Dave Haynie complaining about 65816's supply issues.

Again, #Metoo R&D = behind the competitors.

For X86 PCs, it's a good thing that Zilog 8000 is incompatible with Z80 since Zilog "ground zeroed" Z80 CP/M into a dead end. Don't disturb when competitors make a mistake.

PS; CP/M-86 is not compatible with CP/M Z80 i.e. missing CPU ISA abstraction layer.


Unlike IBM PC, CBM-II doesn't have modular expansion slots, hence it has frozen-in-time graphics capability. The lack of modular partition graphics with C= platforms would act as a deterrent to the platform's evolution and place the platform's graphics upgrade behavior like a game console's hard next-generation switch.

Laptop PCs restored PC's modular partition graphics via USB4's Thunderbolt PCIe support standard.

Last edited by Hammer on 22-Jan-2025 at 12:57 AM.
Last edited by Hammer on 21-Jan-2025 at 02:56 AM.
Last edited by Hammer on 21-Jan-2025 at 02:44 AM.
Last edited by Hammer on 21-Jan-2025 at 02:41 AM.
Last edited by Hammer on 21-Jan-2025 at 02:24 AM.

_________________
Amiga 1200 (rev 1D1, KS 3.2, PiStorm32/RPi CM4/Emu68)
Amiga 500 (rev 6A, ECS, KS 3.2, PiStorm/RPi 4B/Emu68)
Ryzen 9 7950X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB

@Hammer

Quote:

IBM's software consulting service can lead to hardware sales. Hardware without business process software is useless for business.

So you're saying Australia was tricked into buying a heck lot of POWER based IBM hardware or something?

_________________
B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC

@kolla

Quote:

kolla wrote: So you're saying Australia was tricked into buying a heck lot of POWER based IBM hardware or something?

Like you wouldn't believe.

My personal experience is with one of the major 4 banks in 2015, when Australia was implementing a rapid payments clearing standard (PayID). A large global consulting firm were responsible for part of the solution, and they decided to use IBM's pSeries (System p). It did cause us some integration pains between their data in big-endian mode, and the rest of the solution in little-endian mode.

IBM win many large contracts here, especially with government. They fail so often to meet timelines and budgets, but because they have very good lawyers, they keep getting away with it.
Well, almost. Not that long ago, the Queensland Government actually put a ban on hiring IBM.

_________________
All the way, with 68k

@kolla

Quote:

kolla wrote: @Hammer So you're saying Australia was tricked into buying a heck lot of POWER based IBM hardware or something?

Software is useless without hardware. IBM also owns IBM Cloud, formerly SoftLayer which has a large X64 server farm.

IBM mostly serves their traditional B2B and government contracts.

I can obtain cost-competitive POWER9 CPUs, but the platform price is not competitive. A CPU is useless without a motherboard.

Last edited by Hammer on 22-Jan-2025 at 12:48 AM.

_________________
Amiga 1200 (rev 1D1, KS 3.2, PiStorm32/RPi CM4/Emu68)
Amiga 500 (rev 6A, ECS, KS 3.2, PiStorm/RPi 4B/Emu68)
Ryzen 9 7950X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB

@kolla

Quote:

kolla wrote: So you're saying Australia was tricked into buying a heck lot of POWER based IBM hardware or something?

https://www.itnews.com.au/news/services-australia-spends-50m-on-ibm-power-hardware-upgrade-582248

Date: Jul 6, 2022

Services Australia will upgrade the Power Systems servers underpinning its OpenStack private cloud under a new $49.9 million contract with IBM.

Services Australia last purchased Power servers as its former self, the Department of Human Services, through a 2017 contract with IBM worth $32.6 million.

The hardware was used to stand up a Power Systems-based private cloud, which also uses x86 hardware.
...

More broadly, IBM’s $480.7 million mainframe support arrangement with the agency is currently slated to expire at the end of June 2023.

In March, Services Australia handed IBM $28.5 million to upgrade the storage infrastructure in its mainframe environment, suggesting the umbrella support arrangement will be extended.




https://www.hpcwire.com/off-the-wire/ibm-announces-725m-quantum-computing-deal-with-australian-government/\

Date: January 12, 2023

IBM Announces $725 million Computing Services Deal with Australian Government

For the major agencies that partner with IBM – the Australian Taxation Office, Department of Defence, Department of Home Affairs and Services Australia


Australia gov renewed contracts with IBM.

In late 1980 and 1990, I remember my Dad's state (NSW) government job had fleets of IBM PS/2s desktops and a few IBM mainframes. Contracts between Australian governments with IBM are nothing new.

The business process software aspect is an important part of driving hardware sales i.e. staring at pretty Amiga 1000 is not productive. Many of my Dad's workmates had C= A2000 as gaming machines with day jobs working with IBM hardware.

Last edited by Hammer on 24-Jan-2025 at 10:36 PM.
Last edited by Hammer on 24-Jan-2025 at 10:31 PM.

_________________
Amiga 1200 (rev 1D1, KS 3.2, PiStorm32/RPi CM4/Emu68)
Amiga 500 (rev 6A, ECS, KS 3.2, PiStorm/RPi 4B/Emu68)
Ryzen 9 7950X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB

Wow, looks like ppcamiga should move to Australia, I'm sure agami's developers can port MUI to AIX

Regardless, I don't get what all this has to do with IBM spinning off their entire x86 hardware production to Lenovo.

Last edited by kolla on 25-Jan-2025 at 11:18 AM.

_________________
B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC

@kolla

Quote:

kolla wrote: Wow, looks like ppcamiga should move to Australia, I'm sure agami's developers can port MUI to AIX Regardless, I don't get what all this has to do with IBM spinning off their entire x86 hardware production to Lenovo.

IBM selling its X86 PC business to Lenovo effectively placed IBM in a pre-1981 PC position. My point, IBM is still large, but not as large as Microsoft, Google, Apple.

IBM mainframe directly competes against Hewlett-Packard Enterprise (HPE, formerly PA-RISC, CRAY, and non-graphics SGI). SGI transferred its graphics design business to NVIDIA.

HP's PC (formerly Compaq PC, HP printers) business separated itself from HPE to form HP Inc.

The major players in the 1980s have their successors in their strong positions in current times, but missing Commodore International.

_________________
Amiga 1200 (rev 1D1, KS 3.2, PiStorm32/RPi CM4/Emu68)
Amiga 500 (rev 6A, ECS, KS 3.2, PiStorm/RPi 4B/Emu68)
Ryzen 9 7950X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB

@Hammer

Quote:

Hammer wrote: Since Jack Tramiel is marketing 65xx CPUs like soda drinks, MOS's 65xx key engineers have exited CGS e.g. 1. Bill Mensch formed the Western Design Center. 2. Rod Orgill joined Hewlett Packard.

MOS was a failed company when Commodore bought it for peanuts. Everything they did with it afterwards was better than what it would have been.

Quote:

Before Acorn's BBC Micro, Acorn Atom used MOS 6502 and this microcomputer was released in 1980. https://en.wikipedia.org/wiki/Acorn_Atom

The Acorn Atom was not very successful - only ~10,000 were sold.

The heart of the Atom was actually the video chip, a Motorola 6847. I used the same chip in my own computer design in 1981, but with a 6802 CPU (which had the advantage of a tristate bus that significantly reduced the chip count). Years later Tandy did the same with their TRS-80 MC-10.

Quote:

https://www.commodore.ca/commodore-history/the-rise-of-mos-technology-the-6502/ According to Bill Mensch, MOS/CSG had a working 12 Mhz 6502 prototype before Motorola released 12 Mhz 68000 in 1983. Jack Tramiel rejected the 12 Mhz 6502 evolution.

A hand-picked prototype sure, but could they produce them in quantity? Apple found out the truth about Mensch's claims with the 65C816, which wouldn't even work reliably at 4MHz.

The 6502 was mired in fraud from the beginning. First they stole some of the design from Motorola. Then at the launch they sold the CPUs out of fishbowls, not telling the audience that only the ones on top actually worked.

Quote:

The founding of AMD and Intel were due to Motorola's corporate behavior being imported into Fairchild Semiconductor.

The founding of Zilog was caused by Intel's corporate behavior.

Quote:

Jack Tramiel didn't leverage C64's economies of scale with incompatible SuperPET.

The SuperPET was a project of University of Waterloo in Canada, which took a standard 8000 series PET and added a 6809 processor. It was designed for teaching programming in various languages at universities, which obviously was not a mass market.

Not sure what this has to do with the C64, which was released in 1982 (the SuperPET was developed in 1981).

Quote:

This Tramiel product segmentation mindset would be applied to C16/Plus4 vs C64, Atari ST vs MegaST

More rambling?

Many computer manufacturers had a 'product segmentation mindset', including Acorn with their BBC vs Electron and Archimedes A4000 vs A3000, Tandy with the Model 16 (68000 based), Color Computer (6809), MC-10 and Model 100, Amstrad with the CPC, PCW and PC clones, Apple with the Apple IIGS vs Macintosh and Newton. Even PC clones had different hardware to suit different market segments.

Quote:

C128 wasn't the new standard C64 since C128's C900 graphics chipset (MOS 8563) wasn't integrated with the C64's VIC-II for cost reduction.

You have no clue. Putting 80 column text into the VIC-II would be a very difficult job. Timing was already tight at 40 columns. It would have taken years to develop an 80 column version, time that wasn't available.

Turns out having two video chips in the C128 was great for software development, as both could be used at the same time. So not only was this solution quicker and easier, it was more useful too!

The C128 was only intended to be a stop-gap until the Amiga gained traction, but was so popular that they continued making it until 1989. The C65 would have replaced it if they could have developed it in time. Unfortunately it took longer than expected (mostly the engineers' fault) by which time the C64 market had died.

By 1991 the entire home computer market was dying, being replaced by PC clones that had long ago become the 'industry standard'. The Amiga was the last to go. And again it wouldn't have mattered what hardware Commodore put in it. PCs had won the format war and nobody wanted to be on the wrong side of that - which makes perfect sense when a computer is just a tool to do a job.

Luckily for us Commodore made reliable machines that have lasted far longer than anyone expected, with hardware and software that is still being developed today. 10 years ago I wouldn't thought it was possible that we would be where we are now.

Quote:

Unlike the C64, the IBM PC's MDA high-resolution text mode is suitable for killer DOS apps like Lotus 123 and Word Perfect.

'Killer' business apps. The C64 was a home computer designed to be cheap enough for the average family to afford and interesting enough to hold the attention of young and old. Business computers were designed to do boring business stuff, and cost wasn't an issue.

Putting MDA in the C64 wouldn't have made it more appealing. I maintained PC-XTs with MDA in the 80's and they sucked. Then I bought an IBM JX (smd version of the PC jr) and after upgrading it was able to enjoy 16 color 320x200 graphics. For a time it actually held my attention more than my A1000! Why? Because I enjoy programming, especially on machines that have been underappreciated.

The JX was interesting because I could do stuff that hadn't been done on it before. If I had a C128 I would have been similarly motivated.


Quote:

VisiCalc died along with platforms such as DRi CP/M, CBM (8bit business microcomputers), and 'etc'. Lotus 123 2.0 supported X87 FPUs.

Boring!

Quote:

CSG/MOS 65xx CPU is a dead end, hence Acorn's 32-bit ARM existence.

Acorn's ARM didn't save them. I have an Archimedes A3000 sitting in the cupboard because it's not interesting enough - a frustrating machine to use, with a poorly designed GUI and half-baked OS. Also the ARM chip isn't that fast, and programming it in asm sucks.

Quote:

Jack Tramiel's SuperPET's 6809 (with 16-bit ALU math) token selection shows the market that the CSG/MOS 65xx CPU is a dead end.

The SuperPET wasn't Jack's design.

Name one computer line from the 70's that wasn't a 'dead end'. This was the period of innovation - nobody was expecting any one architecture to dominate.

Quote:

C128's Z80 CP/M support in 1985 was overshadowed by DRi's GEM (graphics GUI)/GEMDOS on Atari ST, Apple Mac 512K (GUI), Microsoft Windows 1.0 (GUI) and Commodore's Amiga 1000 (GUI). 1985 for Apple Mac 512K was about next-gen GUI killer app releases.

CP/M was just for legacy support. The CPC also used it, and nobody complained. The PCW used it as the OS and was very successful.

Quote:

During 1984-1985, ARM, 80386, and 68020 were 32-bit CPU releases while Western Design Center released 16-bit ALU-equipped WDC65816.

Blah blah blah. Do you have a point or is this just more rambling?

Quote:

#Metoo R&D = behind the competitors.

Sony waited until 1995 to release a gaming console, and when they did...
Smartphones had been around for years before Apple introduced the iPhone, and then...
VHS came out after BetaMax and was a #metoo design, yet...

Being first doesn't mean your system is best. Most technology is developed over time, and the major players are mostly #metoo. That's how the free market works. We wouldn't have what we do today if all potential competitors just said OK, you go with it and we'll find something else to do.

@Hammer

Quote:

Hammer wrote: Jack Tramiel didn't leverage C64's economies of scale with incompatible SuperPET. This Tramiel product segmentation mindset would be applied to C16/Plus4 vs C64, Atari ST vs MegaST (with blitter in late 1986). With collapsing sales, Tramiel's Atari STE (with blitter) response in 1989 was too late.

What "segmentation" are you talking bout?

Atari's blitter just wasn't ready for the basic ST line and a bit too late for the MegaST launch (but it could be easily added to those that didn't ship with one).
Everything from that point, STe, MegaSTe and Falcon and TT had a blitter.
But in the end the Blitter was about as consequential as ECS was for Amiga (read very limited exclusive SW).
The Falcon was similar to AGA Amiga and the TT the only real oddball.

All in all compatibility was very good, on par with A1000-4000 and surely better than within PCs of that area.
Which lead to more SW banging the HW and even the ROM resulting in inferior compatibility from a user's perspective.

_________________
- We don't need good ideas, we haven't run out on bad ones yet
- blame Canada

@Kronos

Quote:

Kronos wrote: @Hammer Quote: Hammer wrote: Jack Tramiel didn't leverage C64's economies of scale with incompatible SuperPET. This Tramiel product segmentation mindset would be applied to C16/Plus4 vs C64, Atari ST vs MegaST (with blitter in late 1986). With collapsing sales, Tramiel's Atari STE (with blitter) response in 1989 was too late. What "segmentation" are you talking bout? Atari's blitter just wasn't ready for the basic ST line and a bit too late for the MegaST launch (but it could be easily added to those that didn't ship with one). Everything from that point, STe, MegaSTe and Falcon and TT had a blitter. But in the end the Blitter was about as consequential as ECS was for Amiga (read very limited exclusive SW). The Falcon was similar to AGA Amiga and the TT the only real oddball. All in all compatibility was very good, on par with A1000-4000 and surely better than within PCs of that area. Which lead to more SW banging the HW and even the ROM resulting in inferior compatibility from a user's perspective.

Mega ST was released in 1986.

https://en.wikipedia.org/wiki/Atari_ST
Atari introduced the ST1 at Comdex in 1986. Renamed to Mega.

Jack Tramiel could have corrected the ST platform's baseline with a Blitter from 1986. Tramiel's mistake benefits the A500.

The hardware feature's usage is based on the install base.

Unlike Atari, Commodore corrected the early NTSC A1000's missing 64-color EHB mode.

----------------
PC VGA sales passed the 50% mark in 1989 and it's modular with PC graphics i.e. earlier PC builds can upgrade to VGA and SVGA.

According to Dataquest November 1989, VGA crossed more than 50 percent market share in 1989 i.e. 56%.
http://bitsavers.trailing-edge.com/components/dataquest/0005190_PC_Graphics_Chip_Sets--Product_Analysis_1989.pdf

Low-End PC Graphics Market Share by Standard Type
Estimated Worldwide History and Forecast

Total low-end PC graphic chipset shipment history and forecast
1987 = 9.2. million, VGA 16.4% market share i.e. 1.5088 million VGA.
1988 = 11.1 million, VGA 34.2% i.e. 3.79 million VGA.
1989 = 13.7 million, VGA 54.6% i.e. 7.67 million VGA.
1990 = 14.3 million, VGA 66.4% i.e. 9.50 million VGA.
1991 = 15.8 million, VGA 76.6% i.e. 12.10 million VGA.
1992 = 16.4 million, VGA 84.2% i.e. 13.81 million VGA.
1993 = 18.3 million, VGA 92.4% i.e. 16.9 million VGA.
--------------

The majority of Amigas are the A500 Rev6A from 1989, covering Amiga's golden years of 1990 and 1991.

Unlike A500 Rev 5, A500 Rev6A's PCB is ECS-ready which needs ECS Denise.

Quote:

The Falcon was similar to AGA Amiga and the TT the only real oddball.

Falcon has 10,000 to 15,000 unit sales

AGA has about 491,300 units.

AGA sales numbers
A1200, 44,000 (the UK has 30,000 during its launch),
A1200, 100,000 (AF50, Sep 1993),
A1200, 170,000 (AF56, Feb 1994),
CD32, 166,000 (Commodore US president, Jan 1994),
7,500 (Germany's A4000/030),
3,800 (Germany's A4000/040),
Total: 491,300 AGA units.

AGA is not Falcon's sales-level failure. Commodore couldn't produce enough AGA machines due to debt issues.

Last edited by Hammer on 27-Jan-2025 at 04:26 AM.
Last edited by Hammer on 26-Jan-2025 at 09:32 PM.

_________________
Amiga 1200 (rev 1D1, KS 3.2, PiStorm32/RPi CM4/Emu68)
Amiga 500 (rev 6A, ECS, KS 3.2, PiStorm/RPi 4B/Emu68)
Ryzen 9 7950X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB

@bhabbott

Quote:

MOS was a failed company when Commodore bought it for peanuts. Everything they did with it afterwards was better than what it would have been.

Commodore Business Machines (CBM), had invested heavily in the calculator market and was also nearly wiped out by TI's entry into the market. A fresh injection of capital saved CBM, and allowed it to invest in company suppliers in order to help ensure their IC supply would not be upset in this fashion again.

CBM's rescue was done by Irving Gould, not by Jack Tramiel.

CBM's purchase of MOS is done via a stock deal. Shareholders of MOS received a 9.4 percent equity stake in CBM.

Jack "marketing" Tramiel managed the 65xx CPU family like a soda can.

Irving Gould spends his own money to build AGA Amigas.

Herni Rubin played politics to remove Thomas Rattigan.

Quote:

The Acorn Atom was not very successful - only ~10,000 were sold. The heart of the Atom was actually the video chip, a Motorola 6847. I used the same chip in my own computer design in 1981, but with a 6802 CPU (which had the advantage of a tristate bus that significantly reduced the chip count). Years later Tandy did the same with their TRS-80 MC-10. .

Again, BBC Micro wasn't Acorn's 1st 6502-based model. Jack "marketing" Tramiel managed the 65xx CPU family like a soda can.


Quote:

A hand-picked prototype sure, but could they produce them in quantity? Apple found out the truth about Mensch's claims with the 65C816, which wouldn't even work reliably at 4MHz. .

Unlike Western Design Center, ARM had VTI (VLSI Technology, Inc) and later, Apple support.

65C816's success was later with SNES's Ricoh 5A22 which has a 65C816 license. There are 48 million SNES. 65C816 wasn't enough and it was augmented by the official RISC RISC-based SuperFX @ 10 Mhz to 20 Mhz accelerator.

SuperFX's development led to the ARC (Argonaut RISC Core) which scaled to 32bit and 64bit (ARCv3) variants. ARC ISA was changed into RISC-V compatibility as ARC-V implementation.


Quote:

The 6502 was mired in fraud from the beginning. First they stole some of the design from Motorola.

For 6502, any fraud claims are not proven in court i.e. prove it in the court of law.

https://www.commodore.ca/commodore-history/the-rise-of-mos-technology-the-6502/

In June of 1975 Motorola realized they had turned their engineers into their competition. Motorola got mad and sued MOS for infringement of 6800 patents. Chuck said “…there was no substance to their claims…� but it scared the old line industry management at Allen-Bradley. “As soon as lawyers got involved, they wanted out.� said Chuck. As a shock to everyone, Allen-Bradley walked away from MOS and basically gave it to the existing MOS management team.

6500’s performed about 4 times the number of calculations a 6800 could.


With Apple Computer, Inc. v. Franklin Computer Corp, Apple has proven Franklin has copied Apple IP in court.

Quote:

The founding of Zilog was caused by Intel's corporate behavior.

The disagreement was a cost issue since Z80 has a lower cost than 8080.


Like Motorola's 68xx to 68000 shift, Zilog made a mistake with Z8000 being incompatible with Z80 which "ground zeroed" the established Z80 CP/M platforms.


Intel continued 8086 into 32-bit implementation i.e. i386 and IA-32.

Zilog's ISA switch mistake is mirrored on Intel's IA-64 Itanium which AMD exploited i.e. X86-64. Intel is guilty of anti-competitive practices by a court of law.



Quote:

The SuperPET was a project of University of Waterloo in Canada, which took a standard 8000 series PET and added a 6809 processor. It was designed for teaching programming in various languages at universities, which obviously was not a mass market. Not sure what this has to do with the C64, which was released in 1982 (the SuperPET was developed in 1981).

My SuperPET example is to disprove your 8-bit ALU-equipped CSG 65xx is enough claim.

Again, 6809 is equipped with 16-bit ALU math. Jack Tramiel gave the go-ahead for SuperPET's production with a 6809 CPU.

My point, Jack Tramiel didn't execute 16bit ALU math equipped 65xx variant R&D, Tramiel rather buys "off-the-shelf"16bit ALU math 6809 or 16bit Z8000 (for C900) components.

Major 65xx system platform integrators like Apple, Tramiel's Atari, and post-Tramiel Commodore jumped ship to other stronger 32-bit road map CPUs i.e. 68K.

For 1981 product, SuperPET lacks PC's partitioned graphics architecture, a dead-end box like a mobile phone.

Quote:

Many computer manufacturers had a 'product segmentation mindset', including Acorn with their BBC vs Electron and Archimedes A4000 vs A3000,

Compared to Commodore's C64 and Amigas, your cited examples are even larger failures.

Your cited examples lack PC's partitioned graphics architecture.

Quote:

Tandy with the Model 16 (68000 based), Color Computer (6809), MC-10 and Model 100, Amstrad with the CPC, PCW and PC clones, Apple with the Apple IIGS vs Macintosh and Newton.

Color Computer's 6809 is a dead end since 68000 ground zeroed it.

Tandy TRS-80 Model II was 8bit Z80 and a dead end by Zilog's incompatible 16bit Z8000.

Tandy TRS-80 Model 16 has 68000 and a memory upgrade card for Xenix 68K or TRSDOS-16. TRS-80 Model 16 still includes Z80. 68000 is incompatible with Z80. CPU ISA switch has effectively "ground zeroed" the Tandy TRS-80 platform. Meanwhile, the X86 PC clone market evolves with faster X86 CPUs which speeds up existing software investments which is self-feeding.

Tandy TRS-80 Model 16's fortunes died along with VisiCalc and MS Multiplan (failed, text UI). Microsoft learns from Multiplan's #metoo text UI failure with GUI MS Excel for Macintosh and Windows. Microsoft plans to beat Lotus 123 text UI establishment via Mac's "next-gen" GUI Excel and Windows GUI plan.

Tandy TRS-80 all model unit sales are 2.4 million.

https://en.wikipedia.org/wiki/TRS-80#modeliii
By 1982, the company admitted—after no software appeared for the Model 16 after five months—that it should have, like Apple, encouraged third-party developers of products like the killer app VisiCalc. (A lengthy 1980 article in a Tandy publication introducing the TRS-80 version of VisiCalc did not mention that the spreadsheet had been available for the Apple II for a year.)


Unlike Tandy's TRS-80 Model 16, Apple's Steve Jobs made sure the 68000-based Mac 512K had next-gen killer GUI apps in 1985 which boomed in 1986.

Staring at 68000 specs is not productive.

Amstrad CPC had 3 million units sold from 1984 to 1990. Amstrad CPC's 8-bit Z80 CPU family was EOL'ed by incompatible 16-bit Z8000 and 32-bit Z80000. Plus range started from 1990 and the 8-bit Z80 CPU is obsolete for desktop micros.

Amstrad CPC Plus's year 1990 graphics chipset upgrade was #metoo response against A500 and it's too late.

For my family's AUD1500 budget and advice from my older cousin (who has a 286-16/SVGA PC clone), my family purchased the Amiga 500 starter kit with 1084S monitor in 1989 ahead of Amstrad CPC, Atari STE, C64c, C128, Commodore Colt (XT clone). My Dad received a nearly 1-year-old IBM PS/2 Model 55SX (386SX-16, slow VGA, 16bit MCA) from work in 1990.

Quote:

Even PC clones had different hardware to suit different market segments.

Most desktop PC clones have partitioned graphics architecture.

Earlier PC/AT and Compaq 386/AT clones can be upgraded with VGA and SVGA clone cards.

In modern times, laptop PCs with USB4 or Thunderbolt have optional GPU upgrades.

Quote:

You have no clue. Putting 80 column text into the VIC-II would be a very difficult job. Timing was already tight at 40 columns. It would have taken years to develop an 80 column version, time that wasn't available.

Wrong. The missing factor is the strategic direction and C= suffered a brain drain after C64's release e.g. it would take 1 year to re-learn SID's design for the C65 upgrade.


Quote:

Turns out having two video chips in the C128 was great for software development, as both could be used at the same time. So not only was this solution quicker and easier, it was more useful too!

That's useless for re-establishing the platform's baseline specifications. C65 project learns from C128's mistakes with integrated VIC-III.

C65 VIC-III's R&D was underground R&D until it was revealed to Herni Rubin at its nearly completed form which caused 1989's AGA's R&D panic start.


Quote:

The C128 was only intended to be a stop-gap until the Amiga gained traction, but was so popular that they continued making it until 1989. The C65 would have replaced it if they could have developed it in time. Unfortunately it took longer than expected (mostly the engineers' fault) by which time the C64 market had died.

C64/C128's 65xx CPU family didn't evolve fast enough into the 32-bit road map. The problem is rooted in C= management's 65xx CPU's weak R&D investments, hence the switch to 32bit programming model 68000 and the full 32bit 68020+ road map.

C65 R&D chipset was underground R&D until it was revealed to Herni Rubin in its nearly completed state. This is due to weak management direction with core graphics R&D.

From Commodore - The Final Years

C65
1985-1987
Commodore’s rabid legion of C64 fans had patiently awaited a true
sequel and had finally gotten one in 1985 in the form of the C128.
However, the original C64 continued to outpace sales of the C128
and many users wanted better graphics and sound, a faster CPU,
and an improved C64 mode that could run existing C64 software.
Within Commodore, most engineers and managers were done with
the C64 line and ready to move on with Amiga products only. But
certain engineers were not yet ready to say goodbye to the C64
lineage.


C65 proposal design was sent to Ted Lenthe (CSG) on October 14, 1987. VIC-III's underground R&D started around November 1987. VIC-III includes an integrated 80-column character mode.

C65 project initially started with 64 colors and increased to 256 color display targets.

From Commodore - The Final Years

in the early 1980s
when three chip guys, Al Charpentier, Robert Yannes, and their boss
Charles Winterble, had proposed the VIC-20 and C64 computers.
Those engineers had developed their system in secret, away from
the system engineering group, who at the time were focused on
higher end, expensive machines.


Around the early 1980s, a disconnection between low-end VIC-20/C64 and high-end CBM machines due to two separate groups i.e.
1. system engineering group,
2. semiconductor group,

The semiconductor group's C64 project would mostly sustain Commodore until Amiga's revenue takeover.

Blame C= management for weak direction and late R&D starts.

From Commodore - The Final Years

DMAgic
Back in October 1989, Paul Lassa had convinced his team to allow
him to incorporate a simple Direct Memory Access Controller (DMAC)
into the C65. When Lassa was pulled away to help finish the Amiga
3000, he wasn’t able to dedicate much time to the endeavor. Once
his A3000 work was completed, Lassa went back to creating his
prototype DMA chip full time.

He also secretly added blitter functions to the chip, unbeknownst to
Porter. “I realized that I couldn't replicate the whole Amiga custom
chipset but I studied the blitter and its functionality and I
determined that I could do maybe seventy percent of what the
Amiga blitter did in this very small design as part of my DMA
controller.�


From studying Amiga chipset designs, C65's blitter function was secretly added by Paul Lassa without Jeff Porter's knowledge.

At the same time as C65's 1990 R&D, Commodore wasted R&D and production funds on 80,000 C64GS units with 20,000 units sold.

C65 project and corporate politics (i.e. better than Amiga graphics) were a major distraction for the AAA project. 1989 AGA project start was the response to the C65 project's corporate politics.

Last edited by Hammer on 27-Jan-2025 at 04:35 AM.
Last edited by Hammer on 27-Jan-2025 at 04:22 AM.
Last edited by Hammer on 27-Jan-2025 at 03:32 AM.
Last edited by Hammer on 27-Jan-2025 at 03:09 AM.
Last edited by Hammer on 26-Jan-2025 at 11:31 PM.
Last edited by Hammer on 26-Jan-2025 at 11:26 PM.
Last edited by Hammer on 26-Jan-2025 at 10:13 PM.

_________________
Amiga 1200 (rev 1D1, KS 3.2, PiStorm32/RPi CM4/Emu68)
Amiga 500 (rev 6A, ECS, KS 3.2, PiStorm/RPi 4B/Emu68)
Ryzen 9 7950X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB

@Hammer

Quote:

Hammer wrote: Jack Tramiel could have corrected the ST platform's baseline with a Blitter from 1986. Tramiel's mistake benefits the A500. The hardware feature's usage is based on the install base.

Sure, adding a blitter to the ST would have required a new motherboard, which they eventually did with the STe.
This move would have really segmented the market (assuming it would have been widely used) leaving the early adopters behind as the chip couldn't be retrofitted.

Quote:

Unlike Atari, Commodore corrected the early NTSC A1000's missing 64-color EHB mode.

C=/Amiga chip team created a new revision of an existing chip that added a feature but most likely didn't increase costs. Hence C= plunked that version into every Amiga.
Quite a difference.

Quote:

The majority of Amigas are the A500 Rev6A from 1989, covering Amiga's golden years of 1990 and 1991. Unlike A500 Rev 5, A500 Rev6A's PCB is ECS-ready which needs ECS Denise.

The part of ECS that really matters was the 1MB Agnus as it removed the need to produce separate NTSC and PAL versions (for 2MB it was the same die bonded differently). ECS Denise wasn't needed for A500 use cases, I'd go sofar as saying it wasn't needed anywhere (it certainly made no sense on the A3000 with it's FF).

Quote:

Falcon has 10,000 to 15,000 unit sales AGA has about 491,300 units.

And what has that to do with the price of fish?

Both should have been compatible if developers had followed the rules (which quite so often they didn't).
Both were too-little-too-late updates for platforms already past their prime and both had disappointing sales leading to the head company closing shop (one voluntary one not so much).

_________________
- We don't need good ideas, we haven't run out on bad ones yet
- blame Canada

@NutsAboutAmiga

Fair point, in my mind I was thinking about graphics and sound be re-targeted to a video and audio card. So the wonderful chipset I loved the Amiga for originally was now a albatros and be ignored in all my day to day work on my 3K. :)

@Matt3k

50% perhaps not used (copper/sprites/disk io/leds/joysticks/zorro),
but you have 50% ciaa/ciab timers, and Paula sound chip, blitter almost impossible to avoid.

Some of the stuff you can catch with MMU, and perhaps, it might be it can be improved over what have done in NallePuh,

* I think timing emulation can be done better.
(PowerPC has low-level timer instructions, I should not need to worry about forbids, as I did when I wrote the code.)

* Blitter implementation is not great, someone with better understanding of
how the line draw works, should have a look at that.

Last edited by NutsAboutAmiga on 27-Jan-2025 at 11:04 PM.
Last edited by NutsAboutAmiga on 27-Jan-2025 at 10:58 PM.
Last edited by NutsAboutAmiga on 27-Jan-2025 at 10:57 PM.
Last edited by NutsAboutAmiga on 27-Jan-2025 at 10:52 PM.

_________________
http://lifeofliveforit.blogspot.no/
Facebook::LiveForIt Software for AmigaOS