Click Here
home features news forums classifieds faqs links search
6090 members 
Amiga Q&A /  Free for All /  Emulation /  Gaming / (Latest Posts)
Login

Nickname

Password

Lost Password?

Don't have an account yet?
Register now!

Support Amigaworld.net
Your support is needed and is appreciated as Amigaworld.net is primarily dependent upon the support of its users.
Donate

Menu
Main sections
» Home
» Features
» News
» Forums
» Classifieds
» Links
» Downloads
Extras
» OS4 Zone
» IRC Network
» AmigaWorld Radio
» Newsfeed
» Top Members
» Amiga Dealers
Information
» About Us
» FAQs
» Advertise
» Polls
» Terms of Service
» Search

IRC Channel
Server: irc.amigaworld.net
Ports: 1024,5555, 6665-6669
SSL port: 6697
Channel: #Amigaworld
Channel Policy and Guidelines

Who's Online
22 crawler(s) on-line.
 95 guest(s) on-line.
 1 member(s) on-line.


 BigD

You are an anonymous user.
Register Now!
 BigD:  5 secs ago
 pixie:  20 mins ago
 DiscreetFX:  30 mins ago
 matthey:  58 mins ago
 agami:  2 hrs 11 mins ago
 Hammer:  2 hrs 36 mins ago
 a3000d/t:  3 hrs 16 mins ago
 emilianojay:  4 hrs 23 mins ago
 CharlotteMurphy:  5 hrs 13 mins ago
 lewac:  5 hrs 20 mins ago

/  Forum Index
   /  Amiga General Chat
      /  what is wrong with 68k
Register To Post

Goto page ( Previous Page 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 Next Page )
PosterThread
Hammer 
Re: what is wrong with 68k
Posted on 17-Dec-2024 10:18:20
#301 ]
Elite Member
Joined: 9-Mar-2003
Posts: 6171
From: Australia

@matthey

Quote:

1970-1975 4004, 8008/8080
1975-1980 6800, 6502, Z80 dominate

Jack's Commodore purchased MOS and the 65xx CPU's evolution was very slow.

Western Design Center wasn't any better which led to a screaming match between Apple executive for Apple IIGS and Bill Mensch. This is 65816's promised clock speed and mass production issues. Similar problem for later 3Ghz PowerPC 970 promise and late delivery schedule issues.

Apple focused on 68K Macs after the incident with Bill Mensch.

Around 1986, Dave Haynie noted supply issues with WDC's 65816.

Quote:

1980-1985 6502 & Z80 low end, 68k workstation & 16/32 bit embedded markets created

Visicalc's declining fortunes were linked to platforms with just Visicalc. Lotus 123 x86 PC was killing Visicalc.

CP/M Z80 platform was a dead end with incompatible Z8000 while Intel evolved X86 with industry-leading backward compatibility.

Quote:

1985-1990 RISC replaces 68k workstation market, rise of 68k desktop & console, 386 game changer

Microsoft has to beat DOS "killer apps" establishment with Mac GUI influence MS Excel and WinWord.

Windows 2.x install base was 2 million sales by the end of 1989, more than a match against the Amiga platform in terms of numbers. Microsoft released WinWord for Windows 2.1.

Pirated Windows 2.x installations led to more than 2 million install base.

1990's Windows 3.0's 10 million sales (along with Excel and WinWord) were the big nuclear bomb against 68K desktops and DOS "killer apps" establishment. MS Office 1.0 was also released in 1990. Wing Commander's 1990 release completes the double banger. Wing Commander was PC's Defender of the Crown moment.

For the OS/2 project, Windows 3.0's strong sales caused the separation of MS and IBM. MS's in-development OS/2 3.0 was renamed into Windows NT. IBM released 32-bit OS/2 2.0 in 1992.


With timed exclusive games (e.g. Batman Pack), A500's European sales are strong during 1990 and 1991. Commodore made insufficient preparation for SNES's 1992 entry into the European market. Commodore wasted the Amiga's 1990 to 1991 sales boom with pre-VGA/pre-SNES Amiga OCS/ECS. Amiga's 1990-1991 sales boom should been 256 color chipset capable.

For low-end 1992 gaming, C= management forgot to order enough AA chips from HP, and 44,000 A1200 was manufactured for Xams Q4 1992 which caused a 1 million A600 production run against SNES. SNES clubed A600 like clubbing a baby seal.

With just 2 million total sales, Atari's TOS 68K platform faded into history e.g. 10,000 to 15,000 Falcon sales.

-----------------
Custom 65818-based SNES has beaten 68000-based Mega Drive on sales numbers.


Quote:

1990-1995 Mot invests 68k profits into PPC, 68k loses desktop but gains in embedded, x86 desktop!

Most of PowerPC 601's core R&D was done by IBM since early RS/6000's PowerPC 601 didn't have Motorola's 60x bus from 88000.

During 1993, Sony and 3DO were making PR noise about texture-mapped 3D games and 68K wouldn't be a part of it.

During CD32's development and a meeting with Commodore, Psygnosis was pushing for stronger CD32 specs at a minimal cost increase and it was rejected. Psygnosis gave hints for an incoming powerful 3D games console. Sony brought Psygnosis in 1993.

Quote:

1995-2000 68k loses console market but gains in embedded, PPC dies on desktop
2000-2005 ARM Thumb-2 with 68k like code density finally replacing 68k old silicon for embedded

2002, Motorola's MMU-less weak IPC 68000-based Dragon Ball VZ lost to ARMv4T e.g. PalmOS 5 for Tungsten T shipped with ARMv4T based Texas Instruments OMAP @ 144 MHz ARM926T CPU.

The fastest DragonBall Super VZ is a 66Mhz model with a weak 10.8 MIPS.

MMU equiped ARMv4T can run full Linux.

ARM9TDMI has 110,000 transistors.
ARM940T has 800,000 transistors.

----------------------------
X86-64v3 and X86-64v4 have another attempt with smart handheld devices via handheld gaming PC's SteamDeck (AMD Van Gogh SoC) and clones which are powered by Intel's Meteor Lake and Lunar Lake SoCs, AMD's Phoenix Point/Hawk Point, and Strix Point SoCs.

Last edited by Hammer on 17-Dec-2024 at 11:31 AM.
Last edited by Hammer on 17-Dec-2024 at 11:06 AM.
Last edited by Hammer on 17-Dec-2024 at 11:03 AM.
Last edited by Hammer on 17-Dec-2024 at 10:52 AM.
Last edited by Hammer on 17-Dec-2024 at 10:28 AM.
Last edited by Hammer on 17-Dec-2024 at 10:26 AM.
Last edited by Hammer on 17-Dec-2024 at 10:19 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

 Status: Offline
Profile     Report this post  
bhabbott 
Re: what is wrong with 68k
Posted on 18-Dec-2024 9:55:32
#302 ]
Cult Member
Joined: 6-Jun-2018
Posts: 509
From: Aotearoa

@Hammer

Quote:

Hammer wrote:

Jack's Commodore purchased MOS and the 65xx CPU's evolution was very slow.

Define 'very slow'.

The C64 had a 1MHz 6502 in 1981. The C128 had a 2MHz 6502 in 1985, 3 years later. But not much software used it. In that 3 years the C64 had become so popular that developers didn't want to ignore it. The C64 continued to sell over a million units per year until the end of the decade. During this time there was no need for Commodore's 6502 to 'evolve'.

The 65CE02 was developed by Commodore in 1988 (3 years after the C128 used a 2MHz 6502). As well as running at up to 10MHz It had extra registers and was more efficient than the 6502, making it up to 25% faster at the same clock speed. A 'System on Chip' version called the 4510 was designed to go in the C65 (which was supposed to be released in 1990). It had the equivalent of two 6526 CIAs and an MMU addressing up to 1MB built in.

The C65 was canned because by the time it was ready in 1991 the C64 scene was dying. Despite the more than 4 times CPU speed improvement, larger memory and improved graphics, the C65 wouldn't be a success because like the C128 it relied on the C64 user base and software library that didn't need the extra power. There was no significant 'evolution' of the 6502 for the same reason sharks didn't evolve after 23 million years ago - it didn't need to.

However a 4MHz 65CE02 was used in Commodore's A2232 serial card for the Amiga, released in 1991. Here the extra power was useful, and there were no issues with legacy software or the user base so it made sense to use the more powerful CPU.

Some people might say that 3 years between CPU models is 'very slow'. I don't. 3 years was barely enough time for developers to realize the full potential of a new CPU. The 4MHz Zilog Z80A was released in 1977. Tandy used it the TRS-80 Model II in 1979. Sinclair used it in the ZX81 in 1981 and ZX Spectrum in 1982. Amstrad used it in the CPC464 in 1984 and the CPC6128 in 1985 (8 years after the 4MHz version was introduced). Though an 8MHz version was available by then, I don't know of any home computer that used it. The extra speed just wasn't needed.

Intel introduced the 8088 in 1979. The first 'turbo' XT clones running at 8MHz appeared in 1983, 4 years after the 8088 was introduced. But many companies (including Commodore) continued to make 4.77MHz XTs for many years after that. Software continued to be written for a 4.77Mhz 8088 because that's what most users had. PCs had 'turbo' buttons to turn the clock down to 4.77MHz for programs that didn't like going faster. 8MHz wasn't actually that much faster anyway when the bus still ran at 4.77Mhz, so it was more a marketing gimmick than 'evolution'.

Calling Commodore's 6502 development 'very slow' is disingenuous when other CPU manufacturers had similar development times and/or the markets weren't keeping up anyway. But even if it was true, as a software developer I would say that was a good thing - because I wanted a large stable user base, not constantly changing specs and a limited market.

Quote:
Western Design Center wasn't any better which led to a screaming match between Apple executive for Apple IIGS and Bill Mensch. This is 65816's promised clock speed and mass production issues.

True. The problem was that layout techniques which worked on a 2MHz 6502 were weren't good enough for a more complex chip running faster. People who spout 'but Moore's law!" don't understand that each increase in transistor density and speed may involve a paradigm shift requiring huge investment. This could involve a whole new factory to produce chips using a new process. These days that means several billion dollars.

Obviously not all companies have the resources to build a new process line. Some will just continue making chips with an older process because they can't afford to upgrade - a situation that definitely applied to Commodore. The only reason they had a chip making factory at all was because they got it cheap.

So your implication that Commodore was somehow at fault for not evolving the 6502 faster is off base. Firstly they had no need to, and secondly it would have cost too much. If they were in the business of selling chips they might have tried harder, but they weren't so why bother? (answer: so you don't get excoriated by angry Amiga fans 35 years later!).

Quote:
Apple focused on 68K Macs after the incident with Bill Mensch.

The 65816 wasn't good enough anyway. Had they chosen it for the Mac it would have been terribly limiting.

Quote:
Windows 2.x install base was 2 million sales by the end of 1989, more than a match against the Amiga platform in terms of numbers. Microsoft released WinWord for Windows 2.1.

More irrelevancy. The number of users had nothing to do with it. The business world choose IBM in 1981. Even if Commodore sold 10 times more Amigas in 1989 than Microsoft sold Windows, the business world still wouldn't be interested in it. This had nothing to do with the CPU either. The Amiga could have had an 80386 in it and it still wouldn't make a difference when it wasn't IBM compatible.

 Status: Offline
Profile     Report this post  
Hammer 
Re: what is wrong with 68k
Posted on 19-Dec-2024 1:52:03
#303 ]
Elite Member
Joined: 9-Mar-2003
Posts: 6171
From: Australia

@bhabbott

Quote:
Define 'very slow'.

Slower than the competition.


Quote:

@bhabbott
The C64 had a 1MHz 6502 in 1981. The C128 had a 2MHz 6502 in 1985, 3 years later. But not much software used it. In that 3 years the C64 had become so popular that developers didn't want to ignore it. The C64 continued to sell over a million units per year until the end of the decade. During this time there was no need for Commodore's 6502 to 'evolve'.

Are you not aware why major desktop microcomputer platform vendors exited the 65xx CPU family for 32-bit programming model 68000 or ARM?

Your defense for unprepared Commodore is a joke. Are you arguing for R&D to be initiated after the competition releases its "kickass" products?

An R&D road map is about planning and preparing for the future.

Due to RISC threats, Compaq co-funded NexGen in 1986. NexGen was founded in 1986 by Thampy Thomas, and is funded by Compaq, ASCII, and Kleiner Perkins.

http://kpolsson.com/micropro/proc1980.htm
1986: NexGen begins work on the design of a fifth-generation x86 processor, called the F86.
4th gen 486 wasn't released in 1986. Compaq has inside road map knowledge with Intel e.g. Compaq is the 1st PC vendor to release a 486-based PC in 1989.

1988: NexGen begins work on the Nx586, an x86-class processor with performance advantages of RISC processing.

For a startup, that's the level of R&D lead time for 1990's Nx586 (RISC86, released in 1994) and Nx686 (RISC86 v2, AMD K6, released in 1997).

"Only the paranoid survive" - Andy Grove, founder and former CEO of Intel.

Intel started Pentium Pro (P6) R&D in 1991 for 1995 release with 1989 released i960CA with dual instruction per cycle superscalar experience. About a 4-year product development cycle from R&D to product release with relevant experience.

With i960CA with dual instruction per cycle superscalar experience, P5 Pentium R&D was concurrent with Pentium Pro P6's R&D.

For Motorola 88000 R&D
1986, Motorola begins work on the 88000 processor.

There's a long lead time for cutting-edge CPU designs. You're not observant of winning CPU development lead times.

You're a naive fool.
-----------------------------------------
C64's sales were keeping up with the IBM PC until around 1985, hint: "32-bits" from 386, ARM, 68000's 32bit/16bit hybrid Macs, and Atari ST. Amiga wasn't a major platform player until 1987.

C64's production significantly dropped around 1987.

https://pegasus3d.com/total_share.html
1984,
C64 = 2,500,000 (8bit ALU)
Apple II = 1,000,000 (8bit ALU)
PC = 2,000,000 (16bit ALU, PC/AT 286 released)
Mac = 375,000 (16bit ALU)
Intel 286 is fully backward compatible with 8086/8088.
Mac's design foundation was largely driven by Steve Jobs. Mac has mostly business customers.

1985
C64 = 2,500,000 (8bit ALU)
Apple II = 900,000 (8bit ALU)
PC = 3,700,000 (16bit ALU, 286 PC clone arrived).
Mac = 200,000 (16bit ALU)
Amiga = 100,000 (16bit ALU)
Atari ST = 50,000 (16bit ALU)
Intel 386 is fully backward compatible with 286/8086/8088.
Compaq, Intel, and Microsoft team up for Compaq 386 and Windows 2.x 386. IBM dictated MS into 16bit 286 OS/2 1.x road map. Wintel+Compaq 386 team backstabs IBM and DR GEM.

Western Design Center takes over 65xx CPU R&D from frozen CSG/MOS.

Due to CSG 65xx's evolution not being competitive, major 65xx desktop vendors (including Commodore) re-start from ground zero with 68000.

For 68000, Motorola designed a large microcode front end for a 32-bit programming model for a 16-bit ALU-based CPU design with full 32-bit CPU hardware promise e.g. 68020 released in 1984. 68020 has a reasonable 0.5 IPC with ADD instructions. 68020 also includes fast barrel shifter hardware. 68020's MUL instructions would be its weakness which is exploited by workstation 3D RISC competitors.

Intel's 32-bit 80386 bundles PMMU and barrel shifter hardware as standard in 1985.

Unlike the X86 clone CPU market, Motorola's 68K licensees weren't able to create independent "second source" 68K cloners.



1986
C64 = 2,500,000 (8bit ALU)
Apple II = 700,000 (8bit ALU, 16bit ALU via 65816 with supply issues)
PC = 5,0200,000 (16bit ALU, 32bit ALU)
Mac = 380,000 (16bit ALU)
Amiga = 200,000 (16bit ALU)
Atari ST = 200,000 (16bit ALU)
Desktop CPUs with at least 16-bit ALU are beating C64.
Apple IIGS with 65816 was released.
Compaq 386 PC was released.
Steve Jobs founded NeXT and started NeXTSTEP R&D which is the foundation for the later "killer OS" MacOS X.
Steve Jobs' 68K Mac foundation is important for gaining a business use base in the multi-millions.
Supply issues with 65816.
Commodore wasted R&D time on the canceled Ranger in 1986. Henri Rubin cancels Ranger.
Disagreements between Henri Rubin and Thomas Rattigan.


1987
C64 = 1,500,000
Apple II = 500,000
PC = 5,950,000
Mac = 550,000 (includes 32bit ALU 68K release)
Amiga = 300,000
Atari ST = 400,000
Desktop CPUs with at least 16-bit ALU are beating C64.
65xx desktops = 1,500,000
65K desktops = 500,000 (for IIGS?)
68k desktops = 1,250,000. Growth phase for 68K desktops.
Compaq reached $1 billion dollar revenue.
Amiga OCS is not ready for productive resolution mode on a large economic scale.
Commodore management and engineers wasted time on Ranger, monochrome ECS Denise vs four-color ECS Denise vs. eight-color ECS Denise.
Henri Rubin allowed the A2024 R&D workaround hack for Amiga OCS.

Commodore didn't have a seamless C64 to Amiga migration path.

1988
C64 = 1,250,000
Apple II = 200,000 (includes 65816)
PC = 11,900,000
Mac = 900,000
Amiga = 400,000 (includes 32bit ALU 68K released)
Atari ST = 360,000
Desktop CPUs with at least 16-bit ALU are beating C64.
68k desktops = 1,660,000 Growth phase for 68K desktops.
65xx desktop = 1,250,000, Decline phase for desktop markets.
65k desktops = 200,000 (for IIGS?), Decline phase for desktop markets.
Compaq reached $2 billion dollar revenue.
8bit ALU 65xx CPU declines for desktop markets.
Microsoft hired DEC VMS experienced Dave Cutler for a 32-bit OS/2 3.0 (Windows NT) project.
Commodore had signed a contract with ADI in 1988 to manufacture 5000 A2024 monitors.


1989
C64 = 1,250,000
Apple II = 200,000 (includes 65816)
PC = 17,550,000
Mac = 1,100,000
Amiga = 600,000
Atari ST = 300,000
Desktop CPUs with at least 16-bit ALU are beating C64.
68K desktops = 2,000,000. Growth phase for 68K desktops.
65xx/65k desktops = 1,450,000. Decline phase for desktop markets.
Compaq reached $3 billion dollar revenue.



1990
C64 = 700,000
Apple II = 100,000 (includes 65816)
PC = 16,838,000
Mac = 1,300,000
Amiga = 750,000
Atari ST = 300,000
Desktop CPUs with at least 16-bit ALU are beating C64.
68K desktops = 2,350,000. Growth phase for 68K desktops.
65xx/65k desktops = 800,000. Decline phase for desktop markets.
After design flaw-induced delays, the custom A2024 monitor's production run is in the 5000-unit range. A lesson for Commodore shouldn't be in the custom monitor business.


1991
C64 = 800,000
Apple II = 100,000 (includes 65816)
PC = 14,399,000
Mac = 2,100,000
Amiga = 1,035,000
Atari ST = 300,000
Desktop CPUs with at least 16-bit ALU are beating C64.
68K desktops = 3,435,000. Growth phase for 68K desktops.
65xx/65K desktops = 900,000. Markets open for Warsaw pact.


Not including NEC's PC-98 which reach 18 million install base before switching to Wintel MPC standard.

Unlike Intel's X86 CPU family, Commodore didn't plan for a 16-bit ALU 65xx CPU hence the move towards 68K, hence 65xx CPU family was a dead end for desktop markets with no future.

https://www.youtube.com/watch?v=P4VBqTViEx4
Steve Jobs on the role of product and marketing people.

Commodore was largely governed by marketing with C64 and 65xx CPUs being like a Coke drink product.

Steve Jobs's Apple survives while Commodore is dead.

Jack Tramiel is marketing with a Coke drink mentality. Jack Tramiel wasted 1986 MegaST's Bilter release by hard product segmentation difference with ST. STE had a Bilter in 1989 which was too late. Jack Tramiel's incompatible C16 vs C64 Jack attack mentality on ST vs MegaST. Tramiel could have corrected the ST platform standard with 1986 Bilter.

Henri Rubin is an old non-digital electric engineer who can't use a desktop computer. Major conflicts between Thomas Rattigan and Henri Rubin.

Bill Sydnes illegally cloned Apple II via Frankin and IBM's extreme cost reduction PCJr which compromised IBM PC software compatibility.


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

Commodore didn't make C128 as the new standard C64 from 1985 and onwards. Commodore continued to sell both products with hard product segmentation.

A person who purchases C64C can't upgrade to C128's high-resolution mode since graphics architecture is not partitioned. C64C is a dead-end like a game console.

For C64's upgrade path, the Amiga and C128 made the upgrade path confusing for the end user. This is the same schizo mistake as Sega's 32X and Saturn. Commodore didn't deliver a seamless upgrade path from the C64 to the Amiga. C64 emulator (GO-64) was found and the stock 68000 wasn't fast enough to emulate C64. Faster 68020 can emulate a usable C64.
A missed opportunity for A1000Plus/A1200's 68EC020 (released in 1991).

The non-partitioned graphics architecture for Commodore platforms has game console behavior. Unlike IBM, Commodore didn't plan for the platform's future evolution.

Dave Haynie's Acutiator partition graphics architecture was repeatedly delayed by management.


Last edited by Hammer on 20-Dec-2024 at 02:01 AM.
Last edited by Hammer on 19-Dec-2024 at 06:10 AM.
Last edited by Hammer on 19-Dec-2024 at 04:57 AM.
Last edited by Hammer on 19-Dec-2024 at 04:41 AM.
Last edited by Hammer on 19-Dec-2024 at 04:19 AM.
Last edited by Hammer on 19-Dec-2024 at 02:34 AM.
Last edited by Hammer on 19-Dec-2024 at 02:09 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

 Status: Offline
Profile     Report this post  
Hammer 
Re: what is wrong with 68k
Posted on 19-Dec-2024 3:05:07
#304 ]
Elite Member
Joined: 9-Mar-2003
Posts: 6171
From: Australia

@bhabbott

Quote:

More irrelevancy. The number of users had nothing to do with it. The business world choose IBM in 1981.


1. With a business user base majority for the Mac platform, Apple survived. Steve Jobs pushed for high graphics resolution for business over color.

2. IBM PC's 8088 CPU has 16-bit ALU while CSG's 6502 has 8-bit ALU. Some PC clones used 8086. NEC PC-98 used full 16-bit 8086.

Intel: 1
CSG: 0

Intel offered 8087 and 287 FPUs,

Intel: 2
CSG: 0

Motorola 68881 was introduced in 1984 which is late.

Intel: 3
Motorola : 0

For 32-bit computing, Intel ensured its 386 CPU was bundled with MMU standards before Motorola's 68551 and 68030. Intel knows 68000/68010 and 68020 are only half-baked solutions for PMMU-enabled 32-bit OS. Intel didn't execute #metoo tactics against Motorola.

Unix's 68K before 68551 MMU and 68030 don't benefit clonable platforms since they require vendor-specific custom MMU.

Besides the BIOS, most of IBM PC's chip parts are Intel's technology and obtainable from other 3rd parties. Read Micheal Dell's Direct From Dell book.

Without the IBM PC clone, another clonable platform would be in its place e.g. licensable PC-98 from NEC powered by Intel's 8086. NEC licenses PC-98 standard including NEC's ÎĽPD7220 display chip. Microsoft has MS-DOS and Windows for the PC-98 standard.

Intel's second-best platform is NEC's PC-98 standard which had an 18 million install base before the Wintel MPC switch. IBM PC clones blocked NEC PC-98 clones.

ASCII Corporation (Japan) and Compaq co-funded NexGen in 1986.

ASCII Corporation (Japan) is a co-developer with Microsoft's earlier 8-bit Basic UI-driven MSX standard.

ASCII Corporation's co-founder, Susumu Furukawa also co-founded Microsoft Japan (known as Microsoft Kabushiki Kaisha) with Blii Gates. Susumu Furukawa was Microsoft Japan's 1st president.

Motorola didn't have a chance when Intel's second major platform was backed by Japan Inc.

NEC, Intel, and Microsoft have undermined IBM's PS/2 and XGA. NEC was part of the Gang of Nine along with Compaq.

With local partnerships, Intel and MS dominated two of the world's strongest economies in the 1980s i.e. USA and Japan.

Wintel's plan is based on two desktop clonable standards i.e. IBM PC (from USA) and NEC PC-98 (from Japan).

Susumu Furukawa is Locutus of Borg for Japan. PC-98 is the backup X86 desktop cloneable platform.

When one Borg Cube is bad enough, you have two Borg Cubes led by Bill Gates.

"Only the paranoid survive" - Intel's the second X86 PC clone standard backup.

Motorola didn't have a chance.

Following the 8-bit MSX cloneable standard, Intel and Microsoft have a plan for a cloneable 16-bit X86 desktop computer.


3. Jack's Commodore executed hard product segmentation between low-volume SuperPET for business and high-volume C64 for games.


For the 1985 release, C128's high-resolution TEXT mode was too late #metoo with a dead-end Z80 DR CP/M. DR has moved to GEM in 1985.

The mass-produced C64 doesn't have the original IBM PC's high-resolution text mode.

Commodore didn't make C128 as the new standard C64 from 1985 and onwards. C128 has costly VIC-IIE and MOS 8563 dual display chips with no single-chip integration.

MOS 8563's R&D was recycled from Jack era low volume C900.

Commodore continues to sell the C64 along with the C128 with hard product segmentation.

SuperPET wasn't compatible with C64.

1980 VIC-20 wasn't compatible with 1980 PET.

SuperPET's secondary 6809 CPU is a joke (dead end) when it's not compatible with 68000/68008. Due to incompatibility, 6800-based platforms couldn't continue into 68000.

With 68000, Motorola ground zeroed their customer's 6800-based platforms.

With Z8000, Zilog ground zeroed their customer's Z80-based platforms.

Intel: 4
Moto: 0
Zilog: 0


68000's 32-bit programming model road map beats 16-bit Z8000. This is a major factor for 68K's desktop rise.

Moto: 1
Zilog: 0


386 has full backward compatibility with 8088/8086 and 286. 386's 1985 release was timely and Intel secured Compaq's and MS's cooperation in 1985. Intel's timely 386 release in 1985 assured PC cloners a road map towards 32-bit computing as a counter against the rising 68000 related desktop computer announcements. Compaq 386 was released in 1986 along with 1987's Windows 2.01 386.

Intel's other timely releases are 1995's Pentium Pro (killed Advanced Computing Environment's MIPS as PC CPU replacement) and 2024's Lunar Lake (counter Qualcomm Elite X).

AMD64 (X86-64)'s release was timely when IBM released the PowerPC 970 for 64-bit desktops.
AMD64 (X86-64)'s release killed Intel's Itanium (based on HP PA-WIDE propaganda).

HP's PA-WIDE group is from HP's PA-RISC and 68K Unix workstation group.

HP PA-WIDE propaganda is based on "leapfrogging" the competition that assumes the competitors wouldn't catch up. Intel fell for HP's PA-WIDE propaganda, but Intel had project Yamhill EMT64 backup i.e. "only the paranoid survive".


4. Lotus 123 and Word Perfect were the killer apps for MS-DOS PCs.

Word Perfect beats IBM Display Write i.e. being IBM didn't guarantee Display Write's success!


5. Being IBM didn't guarantee PCJr's success since Bill Sydnes had cost-reduced PCJr which compromised PC's "killer apps" software compatibility.


6. Apple killed Apple II clones e.g. Bill Sydnes's illegal Frankin Apple II clone.

IBM couldn't kill PC clones.


7. IBM couldn't kill a Compaq-led gang of nine PC clones. In 1986, Compaq and Western Digital established the IDE standard for PC clones. PC clones followed Compaq's leadership.

8. Being IBM didn't guarantee success since the PS/2 standard lost to Compaq's AT 386 extension standard.

Since 1986, Microsoft, Intel, and Compaq (with the gang of nine PC clones) kicked out IBM from the PC.

Compaq reached $1 billion dollar revenue in 1987.
Compaq reached $2 billion dollar revenue in 1988.
Compaq reached $3 billion dollar revenue in 1989.

Compaq is strong enough to co-fund NexGen X86 CPU clone R&D which AMD purchased in 1996. NexGen leadership dominated AMD's K6 and K7 Athlon R&D.

Compaq's MIPS CPU switch threat was the motivation for Intel's rapid P5, P6, and P7 developments.


9. Being IBM didn't guarantee success for PowerPC.


Quote:

@bhabbott

Even if Commodore sold 10 times more Amigas in 1989 than Microsoft sold Windows, the business world still wouldn't be interested in it. This had nothing to do with the CPU either. The Amiga could have had an 80386 in it and it still wouldn't make a difference when it wasn't IBM compatible.

That's fiction. Amiga OCS was NOT even ready for Mac's high graphics resolution for business.

With a business user base majority for the Mac platform, Apple survived. Steve Jobs pushed for high graphics resolution for business over color. Apple started color Mac R&D in 1985.

Your "Even if Commodore sold 10 times more Amigas in 1989 than Microsoft sold Windows" is irrelevant fiction.

Last edited by Hammer on 26-Dec-2024 at 03:09 AM.
Last edited by Hammer on 19-Dec-2024 at 07:46 AM.
Last edited by Hammer on 19-Dec-2024 at 07:39 AM.
Last edited by Hammer on 19-Dec-2024 at 04:06 AM.
Last edited by Hammer on 19-Dec-2024 at 03:33 AM.
Last edited by Hammer on 19-Dec-2024 at 03:31 AM.
Last edited by Hammer on 19-Dec-2024 at 03:19 AM.
Last edited by Hammer on 19-Dec-2024 at 03:11 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

 Status: Offline
Profile     Report this post  
matthey 
Re: what is wrong with 68k
Posted on 19-Dec-2024 20:21:23
#305 ]
Elite Member
Joined: 14-Mar-2007
Posts: 2456
From: Kansas

kolla Quote:

This nonsense popped up in the late 90s and doesn’t make any sense - IBM had nothing to do with ARexx, so why would CBM license REXX from IBM? Btw, the “full” myth is that IBM got “the looks” of OS 2.x in return from CBM to use with OS/2, which also doesn’t make any sense as CBM got that look from (as you point out) NeXT as well as Motif. There used to be three articles around claiming this CBM/IBM nonsense, all pointing at each other as source.


I believe cross licensing makes sense and is a major reason why businesses create IP banks. Apple sued Microsoft and Xerox sued Apple over the GUI.

https://en.wikipedia.org/wiki/Apple_Computer,_Inc._v._Microsoft_Corp. Quote:

Apple Computer, Inc. v. Microsoft Corporation, 35 F.3d 1435 (9th Cir. 1994), was a copyright infringement lawsuit in which Apple Computer, Inc. (now Apple Inc.) sought to prevent Microsoft and Hewlett-Packard from using visual graphical user interface (GUI) elements that were similar to those in Apple's Lisa and Macintosh operating systems. The court ruled that, "Apple cannot get patent-like protection for the idea of a graphical user interface, or the idea of a desktop metaphor (under copyright law)...". In the midst of the Apple v. Microsoft lawsuit, Xerox also sued Apple alleging that Mac's GUI was heavily based on Xerox's. The district court dismissed Xerox's claims without addressing whether Apple's GUI infringed Xerox's. Apple lost all claims in the Microsoft suit except for the ruling that the trash can icon and folder icons from Hewlett-Packard's NewWave windows application were infringing. The lawsuit was filed in 1988 and lasted four years; the decision was affirmed on appeal in 1994, and Apple's appeal to the U.S. Supreme Court was denied.


Apple and Xerox lost but these kinds of lawsuits are costly and best avoided. Sometimes these cases and legal shenanigans unfairly "win" like the crazy XOR patent that kept the CD32 out the US. Small financially unhealthy businesses can be nuked from orbit by larger financially healthy businesses. Lawfare is not lawfair but a real thing.

I tend to believe that IBM and Commodore cross licensed Rexx and GUI IP. The history is being lost as the Amiga dies but there are non-Amiga sites saying the same thing.

https://www.landley.net/history/mirror/os2/OS2Warp.html Quote:

With Microsoft no longer doing development on the user interface, IBM was faced with creating this themselves. In this timeframe, a deal was made with Commodore. Commodore licensed IBM's REXX scripting language for inclusion in their AmigaOS, and IBM took many GUI design ideas from the AmigaOS for their new GUI. With the release of OS/2 2.0, the WorkPlace Shell (WPS) user interface was born. OS/2 was now a 32-bit operating system, with a fully object-oriented graphical user interface. Based on IBM's System Object Model (SOM), the WorkPlace Shell is still the model for all graphical user interfaces, since nothing else has come even close to providing the same functionality. OS/2 2.1 and 2.11 followed, including a version of 2.11 with full Symmetric Multi-Processing (SMP) support, the first desktop operating system to support multiple CPUs. OS/2 2.x won over many Windows 3.x users because of it's ability to run Windows programs seamlessly, while maintaining a stable system, something that Windows had trouble doing. IBM even went so far as to trademark the term "Crash-Proof."


The site is an old OS/2 fan site with some OS/2 propaganda and misinformation but mostly good info. There is a link to a timeline for OS/2 written by David Both on the same site.

https://www.landley.net/history/mirror/os2/OS2History.html Quote:

By late 1990, Microsoft had intensified its disagreements with IBM to the point where IBM decided that it would have to take some overt action to ensure that OS/2 development continued at a reasonable pace. IBM, therefore, took over complete development responsibility for OS/2 1.x, even though it was in its dying days, and OS/2 2.00. Microsoft would continue development on Windows and OS/2 3.00. Shortly after this split, Microsoft renamed OS/2 V3 to Windows NT.


OS/2 history is fading too. There is some OS/2 history I found interesting and did not know. It sounds like Microsoft ripped off IBM and OS/2 to created Windows NT, which is what Windows is today. There is misinformation though too.

https://www.landley.net/history/mirror/os2/OS2History.html Quote:

OS/2 2.00 - 1992

OS/2 2.00 was released in the spring of 1992. The first true 32 bit operating system for personal computers (and for years the only one), it met IBM's stated goal of being a better DOS than DOS and a better Windows than Windows. It did this through the use of Virtual DOS Machines (VDMs) which allowed OS/2 to run many DOS (and Windows) programs at the same time as though they were on completely separate computers. As far as the DOS programs were concerned, they actually were in separate computers. Windows programs run on IBM's licensed version of Windows 3.1 called Win-OS/2.


I would certainly argue that the original AmigaOS for the 68000 is 32-bit because the ISA and pointers are 32-bit giving a large flat 32-bit address space. That likely means Sinclair QDOS for the 68008 Sinclair QL was not only the first preemptive multitasking OS for a PC but the first 32-bit OS for a PC. Google search AI is wrong.

Google: What is the first 32-bit OS for a personal computer?
AI Overview: The first 32-bit operating system for a personal computer was Windows NT developed by Microsoft, specifically the version "Windows NT 3.1" which was released in 1993; it was considered the first fully 32-bit version of Windows.

Google: What is the first preemptive multitasking personal computer?
AI Overview: Microware's OS-9 was one of the first preemptive multitasking operating systems (OS) available to home users. It was available for computers based on the Motorola 6809, such as the TRS-80 Color Computer 2.

Google: What is the first personal computer that came with preemptive multitasking?
AI: Preemptive multitasking was implemented in the PDP-6 Monitor and Multics in 1964, in OS/360 MFT in 1967, and in Unix in 1969, and was available in some operating systems for computers as small as DEC's PDP-8; it is a core feature of all Unix-like operating systems, such as Linux, Solaris and BSD with its derivatives, ...

Google: What is the first personal computer that came with multitasking?
AI Overview: The Amiga was one of the first personal computers to have a multitasking operating system available to consumers:
AmigaOS
Commodore International developed AmigaOS, which was introduced in 1985 with the Amiga 1000. It was a single-user multitasking operating system that included a built-in animation system and four channels of sampled sound.

Does AI spread more misinformation than old websites? Do the winners write the history books while much of the other history is lost?

Last edited by matthey on 20-Dec-2024 at 01:54 AM.
Last edited by matthey on 19-Dec-2024 at 08:24 PM.
Last edited by matthey on 19-Dec-2024 at 08:24 PM.

 Status: Offline
Profile     Report this post  
Hammer 
Re: what is wrong with 68k
Posted on 20-Dec-2024 1:04:34
#306 ]
Elite Member
Joined: 9-Mar-2003
Posts: 6171
From: Australia

@matthey

Quote:
Apple and Xerox lost but these kinds of lawsuits are costly and best avoided. Sometimes these cases and legal shenanigans unfairly "win" like the crazy XOR patent that kept the CD32 out the US. Small financially unhealthy businesses can be nuked from orbit by larger financially healthy businesses. Lawfare is not lawfair but a real thing.

I tend to believe that IBM and Commodore cross licensed Rexx and GUI IP. The history is being lost as the Amiga dies but there are non-Amiga sites saying the same thing.


From Commodore - The Final Years

New UI Look

(SKIP)

Much like the hardware team, the software engineers started
projects entirely on their own. “We were all kind of creative people so
we always needed our creative outlets to have a little fun or do
something different or add our own little footprint to things,” says Eric
Cotton. “We used our own experience and our own talents and we
were even influenced by some of our video game backgrounds.”
At the time, graphical user interfaces on the Amiga, Atari ST, and
Macintosh looked rather flat. “We were keeping our eyes on what
was going on around us,” says Cotton. In late 1988, Commodore
acquired a curious new computer. “We actually had a NeXT computer
that Commodore bought to examine.”


The NeXT was the brainchild of Steve Jobs, who had a particular
talent for taking existing innovations and wrapping them up into a
harmonious whole. Jobs stole the look of his NeXTSTEP operating
system from the Silicon Graphics IRIX 3 operating system.
The
interface looked almost 3D, as though it had depth on the flat screen.
Although the NeXT was a commercial failure at $6,500, it influenced
other computer makers and operating systems of the time.


--------------------------
For NeXTStep, NeXT steals the 3D GUI look from SGI's IRIX 3
For Workbench 2.x, C= steals the 3D GUI look from NeXT.
For Windows 95, MS steals the 3D GUI look from NeXT.

Steve Jobs famously said in 1996: "Picasso had a saying -- 'good artists copy; great artists steal' -- and we have always been shameless about stealing great ideas."

Advanced Computing Environment (ACE)'s MIPS R-series CPU selection dominance was influenced by SGI. ACE is the next non-x86 PC clone standard.

Windows NT 3.51 supported OpenGL ICD.
Amiga Hombre targeted OpenGL.

The road map leader is SGI.

Intel and Microsoft would be chasing down most feature lists from the SGI workstation.


The OpenGL standard is now defined by an Architecture Review Board (ARB) consisting of Digital Equipment Corporation, Evans & Sutherland, IBM, Intel, Intergraph, Microsoft, and Silicon Graphics.

Microsoft and Silicon Graphics have worked together since 1991 in the development of OpenGL for use with the Windows NT operating system. With Windows NT Workstation version 3.51, Microsoft has released the second version of OpenGL for this operating system


PS; My dream machine was an SGI 3D graphics workstation.

MIPS R8000 (1994) was the first superscalar MIPS design, able to execute two integer or floating point and two memory instructions per cycle. The R8000 can dispatch up to four instructions including two memory accesses per cycle.

1995's Pentium Pro P6 (up to 3 fused instructions per cycle) countered MIPS R8000, hence killing Advanced Computing Environment's RISC-based MIPS PC clone initiative.

Nintendo and Sony continue with the MIPS R-series for the games console market segment until IBM's PowerPC.

Last edited by Hammer on 20-Dec-2024 at 01:15 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

 Status: Offline
Profile     Report this post  
bhabbott 
Re: what is wrong with 68k
Posted on 20-Dec-2024 2:55:50
#307 ]
Cult Member
Joined: 6-Jun-2018
Posts: 509
From: Aotearoa

@matthey

Quote:

matthey wrote:

Does AI spread more misinformation than old websites?

Yes. Google's AI is not trustworthy, it will tell you things that are not true because the context missing, or because it has mixed stuff up. This not surprising given how it works.

Old websites may be wrong, but Google's AI makes it worse. If a link is provided always check it out - much of the time it does not say what the AI implies. If there isn't a link then just ignore it because there's no way to tell how accurate it is.

Quote:
Do the winners write the history books while much of the other history is lost?

Yes. Or more accurately the survivors write the history books.

Brian Bagnall's book are good because he has collected information about Commodore from all over, and interviewed many of the 'survivors' to provide a more balanced record. He also refrains from slanting it to support a narrative, which is a rare thing in Amiga Land.


 Status: Offline
Profile     Report this post  
Hammer 
Re: what is wrong with 68k
Posted on 20-Dec-2024 3:37:26
#308 ]
Elite Member
Joined: 9-Mar-2003
Posts: 6171
From: Australia

@bhabbott

Quote:

Even if Commodore sold 10 times more Amigas in 1989 than Microsoft sold Windows, the business world still wouldn't be interested in it. This had nothing to do with the CPU either. The Amiga could have had an 80386 in it and it still wouldn't make a difference when it wasn't IBM compatible.


For 68K's 3rd best desktop platform, Atari ST platform with Jack Tramiel mistake example,

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

3D computer graphics applications (like Cyber Studio CAD-3D, which author Tom Hudson later developed into Autodesk 3D Studio), brought 3D modelling, sculpting, scripting, and computer animation to the desktop. Video capture and editing applications use dongles connected to the cartridge port for low frame rate, mainly silent and monochrome, but progressed to sound and basic color in still frames. At the end, Spectrum 512 and CAD-3D teamed up to produce realistic 512-color textured 3D renderings, but processing was slow, and Atari's failure to deliver a machine with a math coprocessor had Hudson and Yost looking towards the PC as the future before a finished product could be delivered to the consumer

Atari's late FPU inclusion led Tom Hudson to jump on the PC with FPU support.

Meanwhile, Apple's Macintosh II supported 68020/68881 @ 16Mhz in 1987 which is one 1 year after Compaq 386 release. The same Compaq 386 can be upgraded with ET3000AX in 1987.

------------------------------
The Amiga had a healthy 3D raytracing software sector since Commodore released A2620 (with 68020/68551/68881) in 1988. Without 68551 MMU being late, A2620 would been released in 1987. Faster 030/882 accelerators followed.

A2000 annual sales are numbered in about 1 :10 ratio (?) with A500.

For 68K 2nd best desktop platform, the Amiga,

Blender 3D's history started with the Amiga. https://zgodzinski.com/blender-prehistory/

Q: When was the Amiga your main development tool?

Ton Roosendaal: 1987 - 1991 (summer).

Ton Roosendaal: I started with an Amiga 500 with 512 Kb. Bought the memory expansion to make it 1 MB.

Then in 1988 we invested in two Amigas 2000 with “turbo cards”. I believe one was with 68020, the other 68030. Memory went up to 8 MB. Hard drives had 40 MB. We backed up everything on floppies first and on Magneto-optical drives after. In the end we also bought a 68040 card for rendering.


Around 1991, the Silicon Graphics workstation (MIPS-based) replaced the A2000.


Q: Why did you abandon Amiga and moved on to Silicon Graphics?

Ton Roosendaal: The main reason was that Amiga itself proved to be a dead-end for production. The market demanded higher quality graphics, especially 24 bits color. If Amiga would have made that step sooner…


The Amiga ECS disappointment.

Independent clone VGA vendors to improve PC graphics performance without IBM's intervention. For PC's chipset improvements, Intel dropped PCI 1.0 (1992) nuke which displaced NEC-led VL-Bus (VESA) and IBM's 32bit MCA. CPU manufacturers like Intel have used their market power to impose PCI i.e. Intel PCI chipset competes in the market with Taiwanese PCI chipset clones following e.g. VIA, SIS, ALI, UMC and 'etc'. Motorola didn't have Intel's desktop chipset market power.

You're not addressing Amiga OCS issues for business use cases.

_________________
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

 Status: Offline
Profile     Report this post  
bhabbott 
Re: what is wrong with 68k
Posted on 20-Dec-2024 3:46:17
#309 ]
Cult Member
Joined: 6-Jun-2018
Posts: 509
From: Aotearoa

@matthey

Quote:

matthey wrote:

The 68000 dominated the 8086.

performance 68000
orthogonality 68000
GP registers 68000
address space 68000
datatypes 68000
code density tie (68k is better than x86 though)
support chips 8086
compatibility 8086
price 8086


Intel was lucky to land the IBM PC contract and they did so by selling at least early 8088 CPUs below cost so the high margin desktop market was far from instantaneous.

Yes, Intel was lucky to have landed the IBM PC contract. Before that the 8086 was struggling, and yes the 68000 did dominate it technically.

But IBM had good reasons to choose the 8088 over the 68000, and once it did it was game over.

Well not quite game over. We have Commodore to thank for taking onboard a dream machine that by rights should have sunk into obscurity. Jack Tramiel probably would still have used the 68000 if he had remained at Commodore, but the machine wouldn't be nearly as good. In many ways the Amiga was a happy accident of history. All the fans who dis it for one thing or another are off base, because it was way better than expected.

In a less tumultuous world we probably would have gotten something like the ST but with VIC and SID. They might even have stuck a 6502 in it to run C64 software, then it would take on the roles of the C128 and C65. Or they might have tried to make a Unix-based business computer like Tandy did, which would fail badly.

There's not much point speculating about what might have been, when most realistic scenarios would have led to a worse result. If Motorola had managed to produce the 68000 earlier and convinced IBM to use it, then goodbye Amiga because there would be no reason not to make it IBM compatible. It would just be another quirky PC clone. I suppose you could say that's good because then the 68000 would dominate the industry, but we would be poorer for it.

Because the Amiga was different, and because it was frozen in time in 1994, it has become a great retro computer worth receiving our love. Compare that to the PC, where machines from that era are meh. 68k being frozen in time a few years later makes it even better. We get to play around with FPGAs and PiStorms etc. as well as vintage 68k CPUs, which wouldn't happen if there was continuous development up to modern times.

 Status: Offline
Profile     Report this post  
bhabbott 
Re: what is wrong with 68k
Posted on 20-Dec-2024 4:01:32
#310 ]
Cult Member
Joined: 6-Jun-2018
Posts: 509
From: Aotearoa

@Hammer

Quote:

Hammer wrote:

Blender 3D's history started with the Amiga. https://zgodzinski.com/blender-prehistory/

Q: When was the Amiga your main development tool?

Ton Roosendaal: 1987 - 1991 (summer).

Ton Roosendaal: I started with an Amiga 500 with 512 Kb. Bought the memory expansion to make it 1 MB.

Then in 1988 we invested in two Amigas 2000 with “turbo cards”. I believe one was with 68020, the other 68030. Memory went up to 8 MB. Hard drives had 40 MB. We backed up everything on floppies first and on Magneto-optical drives after. In the end we also bought a 68040 card for rendering.


Around 1991, the Silicon Graphics workstation (MIPS-based) replaced the A2000.


Q: Why did you abandon Amiga and moved on to Silicon Graphics?

Ton Roosendaal: The main reason was that Amiga itself proved to be a dead-end for production. The market demanded higher quality graphics, especially 24 bits color. If Amiga would have made that step sooner…

This is so irrelevant it's ridiculous.

Person starts with a 'toy' A500, then decides they need a high-end graphic workstation costing almost as much as a house - and this somehow proves that there was something wrong with the Amiga?

No Ton, Commodore shouldn't have made that step just for you. Go buy your Silicon Graphics thing and leave us to enjoy our computers for the masses, not the classes.

 Status: Offline
Profile     Report this post  
bhabbott 
Re: what is wrong with 68k
Posted on 20-Dec-2024 4:21:10
#311 ]
Cult Member
Joined: 6-Jun-2018
Posts: 509
From: Aotearoa

@Hammer

Quote:

Hammer wrote:

You're not addressing Amiga OCS issues for business use cases.

No need, because there weren't any issues worth worrying about.

Despite not being primarily a 'business' machine, the Amiga had plenty of 'business' software that ran fine in OCS. We used an A2000 to make all the professionally offset-printed manuals, box artwork and CD inlays for the titles we produced. It worked very well. I and my friend both ran our businesses on the Easy Ledgers accounting package, which works perfectly in OCS. I also used the Professional Calc spreadsheet program, and GP-Fax (which we used for all our faxing needs) both of which worked fine in OCS.

I'm struggling to think of any typical business application in the late 80's to early 90's that OCS wouldn't be able to handle.

 Status: Offline
Profile     Report this post  
Hammer 
Re: what is wrong with 68k
Posted on 20-Dec-2024 5:25:59
#312 ]
Elite Member
Joined: 9-Mar-2003
Posts: 6171
From: Australia

@bhabbott

Quote:
This is so irrelevant it's ridiculous.

Person starts with a 'toy' A500, then decides they need a high-end graphic workstation costing almost as much as a house - and this somehow proves that there was something wrong with the Amiga?

No Ton, Commodore shouldn't have made that step just for you. Go buy your Silicon Graphics thing and leave us to enjoy our computers for the masses, not the classes.


Are you arguing 24-bit display is limited to SGI? Hint: PC's SVGA cards can display 24-bit color (true color) on various models.

Commodore - The Final Years

Porter specifically wanted 1000 by 800 resolution with 8 bit planes and 16 million colors—something to exceed the current competition.

Of importance would be keeping the new video technology compatible with existing commercial Amiga software.

All of these plans would be discussed at Commodore’s worldwide engineering meeting, scheduled for September 22, 1987 at the Embassy Suites hotel in New York. The meeting was called by Henri Rubin, and those invited included the main West Chester engineers, along with engineers from Germany and Japan.


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

1. 1000 by 800 resolution, 8 bitplans and 16 million colors. AGA has 8 bitplanes and 18-bit HAM8 (lossy color compression) display capability.

2. Backward compatible.

3. Jeff Porter's advocacy before September 1987. Dale Luck advocated for a 16-bit chunky display, which is similar to 3DO's 16-bit chunky color sustained display performance mode.

4. Committee decision-making pace.

For CPU power beyond 68040 and 80486, the PC was a little late when Pentium was demo'ed in 1992 and released in 1993. Pentium's 1992 demo and 486 overdrive socket assure PC cloners and customers there is an upgrade path. Intel demo'ed 486DX at 100Mhz in 1992. These are defensive tactics from Intel.

For memory-protected OS, Windows NT 3.1 was released in 1993, hence MS was a little late in this area.

SGI Indigo's display options come in
8-bit entry graphics,
24-bit color XS graphics,
XZ graphics with double the geometry performance over XS,
Elan Graphics with four times geometry performance over XS.

SGI Indigo's CPU range from MIPS R3000A @ 33Mhz, R4000 @ 100 Mhz (after Aug 1991), and R4400 @ 150Mhz (after Nov 1992).

PS1 has a cost-reduced MIPS R3000A-based CPU @ 33 Mhz with GTE for geometry @ 58 Mhz

Meanwhile, around 1991, the multimedia group (Jeff Porter) selected MIPS-X-based CL-450 @ 40Mhz for CDTV-CR's FMV in parallel to Jeff Frank's A600 R&D. CL-450-based FMV has a 24-bit display chip.

68060 @ 50Mhz was released in April 1994.

Intel would rapidly crank up Pentium's clock speed to 66 Mhz in March 1993, 100 Mhz in March 1994, and 150 Mhz Pentium Pro in 1995.

SGI's first R8000 with 75Mhz was introduced in June 1994.
SGI released the R8000 with 90 Mhz in mid-1995. SGI struggled with ramping up clock speed.

PowerPC camp largely has kept up with Intel's clock speed ramping until the Intel Pentium MMX/Pentium II vs AMD K6 MMX/K6-II 3DNow Mhz and Pentium III vs Athlon Ghz race.

Pentium III vs Athlon Ghz race dethrones Mhz king Alpha CPU family.

K8 Athlon defeated PA-RISC's PA-WIDE-based Itanium successor.

After ARM defeated MIPS in the mobile space, ARM stayed in the shadows in its mobile safe space until it re-engaged the desktop market during a renewed Intel vs AMD core count and Ghz race.

From the US government-funded US academia, RISC-V replaced MIPS with the lessons from ARM vs MIPS. US government-funded US academia and DARPA-supported MIPS.

Japan and UK governments support ARM. There are national security issues with CPUs.

Last edited by Hammer on 20-Dec-2024 at 11:07 AM.
Last edited by Hammer on 20-Dec-2024 at 05:34 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

 Status: Offline
Profile     Report this post  
Hammer 
Re: what is wrong with 68k
Posted on 20-Dec-2024 6:41:21
#313 ]
Elite Member
Joined: 9-Mar-2003
Posts: 6171
From: Australia

@bhabbott

Quote:

No need, because there weren't any issues worth worrying about.



Are you aware that CDTV-CR's MPEG1 FMV module has a 3rd party 24-bit color display chip?

From Commodore - The Final Years

Unfortunately, the final cost of the CDTV-CR came in higher than
Mehdi Ali had anticipated. Porter had hoped for under $300 as his
low-ball estimate, but the final cost was $351.19. “What ended up
happening is after it gets built, it's a really nice machine and it's
really slick and they've also incorporated the potential for doing
MPEG decoding, so it can play movies and that kind of thing,” says
Carl Sassenrath. “And it ends up not meeting the cost estimates. So
it was a little more expensive.”

CDTV-CR's $351.19 cost includes an FMV module that was recycled for CD32.

Wasted FMV module's compute power and 24-bit color display capability for playing VCD.

AGA wasn't optimized to display VCD's 24-bit color when 3rd party display chip can do it!

CDTV-CR with MPEG1 FMV is like A600 with half of PS1. CL-450 SoC-based FMV is fully contained SBC with custom MIPS-X RISC CPU @ 40Mhz and custom MPEG-related extensions.

US government DARPA supported MIPS-X CPU design.

http://i.stanford.edu/pub/cstr/reports/csl/tr/86/289/CSL-TR-86-289.pdf
DARPA MIPS-X has 32bit 32 GPR.

Customized CL-450's MIPS-X has 32 24-bit GPR.

Quote:

@bhabbott

Despite not being primarily a 'business' machine, the Amiga had plenty of 'business' software that ran fine in OCS.

Not good enough for mainstream DTP markets and other productive apps that need stable high-resolution productivity modes.

Macs defeated Amiga OCS in publishing and back-office markets.

Quote:

@bhabbott

We used an A2000 to make all the professionally offset-printed manuals, box artwork and CD inlays for the titles we produced.

Our non-profit publishing production (e.g. printed materials) uses Macs. I was able to use my A3000 as a Mac LC clone.

Amiga's Deluxe Music 2.0 is also used for the non-profit organization and I used IntroCAD and X-CAD 2000 for my school work.

A3000's Amber chip with VGA monitor was okay, it needs to be cheaper, mass-produced, and higher color display. ECS wasted the faster 32-bit Chip RAM (7.1 MB/s via CPU access cycles similar to AGA machines).

Between OCS vs ECS, I'm not going back to OCS e.g. I upgraded my 2nd A500 rev6A's OCS Denise to ECS Denise. My Dad traded 1st A500 rev6A for ex-corporate A3000. I traded A3000 for a Pentium 166-based PC in 1996.

ECS Denise with ECS Agnus also improves Paula's audio output to 56 kHz.

I lug A3000 for school for presentation work. A3000 doesn't need a special monitor since the school's PC SVGA monitor and composite video out (for large CRT TV) would do the job. I have shown A3000's fake Mac mode to my chemistry/biology teachers. CanDo and Scala 500 are used for presentation.

A1200 would be my favorite Amiga since it's semi-portable with AGA graphics capabilities. EC020 @ 25Mhz and 2MB Fast RAM would make it better for small price increases.

If the A1200 had been brought instead of my A3000, I would have kept the A1200. I prefer AA1000Plus with two Zorro II slots plans.

After AGA and ECS, I wouldn't go back to OCS.

My Dad's work friends have A2000s and they mostly play games on it i.e. the Amiga was the gaming PC.

Quote:

@bhabbott

It worked very well. I and my friend both ran our businesses on the Easy Ledgers accounting package,

Did you assume I wasn't aware of Easy Ledgers?

Due to the lower resolution of the Amiga OCS, 1989 Easy Ledgers' reports formatting and presentation were inferior to Mac's Quicken Quickbooks. Easy Ledgers A6's reporting presentation looks like a DOS text version. WTF?

I reviewed Easy Ledgers rev A6 (1989) ). I rather use A3000's fake Mac to run Quicken Quickbooks. Easy Ledgers 2.0 was released in 1994.

With a fake Mac, the A3000/030/882@ 25Mhz can run Quickbooks, MS Excel, MS Word, Mathlab and etc, hence A3000 with VGA monitor redeems itself when it sits next to a Windows 3.1/386DX-33/ET4000AX PC. A3000 just needs a higher color display.

A3000 with ShapeShifter (fake Mac) double duties for school and non-profit organization work. A3000 can still play A500 class games with a degrader tool and soft Kickstart 1.3.

Both my A3000 (faking a Mac LC) and Windows 3.1/386DX-33/ET4000AX-based PC runs a similar set of applications i.e. Quickbooks, MS Excel, and MS Word. PC has additional applications from government work e.g. Ami Pro. The apps are from work.

A500/A2000 must have ECS in 1988. AGA in 1991 (1 year behind SNES is maximum).


Quote:

@bhabbott

which works perfectly in OCS. I also used the Professional Calc spreadsheet program, and GP-Fax (which we used for all our faxing needs) both of which worked fine in OCS.

Did you assume I didn't review Professional Calc (v1.0 released in 1991)? Many Amiga's applications are on British magazine cover disks.

British magazines attempted to move the Amiga platform from games into business e.g. Pro Calc V2 full version is included with CU Amiga Cover Disk 94.

Professional Calc 2.1 (1993) is optimized for the 640x400p/480p/512p line since the 640x256p resolution has a squashed look.

https://archive.org/details/microsoft-excel-for-macintoshtm-3.0-1991
MS Excel For Macintosh 3.0 (1991)

From https://www.youtube.com/watch?v=2UYTZNllCaM
Spreadsheets on the Amiga
Amiga's spreadsheets in the 80s were close to PC DOS counterparts. They started to smart up their GUI designs starting with WB2.0.

Both MS Excel 3.0 and Professional Calc v1.0 are year 1991 releases.

TurboCalc v2.05D is a 1993 release.

For 68K platform, there are reasons for Mac defeated Amiga in the business market.

MS Excel GUI established a sizeable install base for GUI Windows 2.x (from 1987) and 3.0 (1990) and MacOS (from 1985) when Lotus 123 for Windows 1.0 was released in 1991. For GUI, MS flips the spreadsheet establishment towards Excel. 1991 released Lotus 123 for Windows 1.0 is fighting from behind.

Apple's strength is GUI craftmanship.

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

A similar situation for SNES vs AGA; when SNES was released on the European market in 1992, SNES had a worldwide install base lead while Commodore struggled with 44,000 A1200 for XMas Q4 1992. This is due to management forgetting to order enough AA chips from HP without factoring lead times which caused Mehdi Ali to order 1 million A600s.

By 1992, PC VGA 256 colors and SNES dominated game artwork design which shows aging OCS/ECS while AGA starting to establish its install base.

Quote:

@bhabbott
I'm struggling to think of any typical business application in the late 80's to early 90's that OCS wouldn't be able to handle.

You wouldn't realize why Amiga was defeated by Macs on publishing and back-office markets.

https://www.youtube.com/watch?v=eywi0h_Y5_U
Ballmer Laughs at the iPhone, Ballmer lists feature bullet points for Windows Mobile. Ballmer didn't factor in GUI craftmanship. Google Android nearly copies iOS.

Ballmer = You. You don't factor in GUI craftmanship.

Amiga Format is created on QuarkXPress on the Mac. Amiga can run QuarkXPress via fake Mac methods.

Last edited by Hammer on 20-Dec-2024 at 10:29 AM.
Last edited by Hammer on 20-Dec-2024 at 09:14 AM.
Last edited by Hammer on 20-Dec-2024 at 08:08 AM.
Last edited by Hammer on 20-Dec-2024 at 07:43 AM.
Last edited by Hammer on 20-Dec-2024 at 07:20 AM.
Last edited by Hammer on 20-Dec-2024 at 07:19 AM.
Last edited by Hammer on 20-Dec-2024 at 07:04 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

 Status: Offline
Profile     Report this post  
OlafS25 
Re: what is wrong with 68k
Posted on 20-Dec-2024 11:14:49
#314 ]
Elite Member
Joined: 12-May-2010
Posts: 6472
From: Unknown

@Hammer

As I wrote in the other discussion too, in 2024 hardware more or less is irrelevant. The only exception is 68k as a retro platform. More important is to define use cases for the platform. OS itself should be portable and we need concepts that offer some sort of future besides being a retro platform for a slowly aging community. People today use devices they buy for a certain task for a certain price. They mostly do not care about the hardware, often not even know it. So more important are use cases for the future.

 Status: Offline
Profile     Report this post  
Hammer 
Re: what is wrong with 68k
Posted on 20-Dec-2024 11:28:55
#315 ]
Elite Member
Joined: 9-Mar-2003
Posts: 6171
From: Australia

@bhabbott

Quote:
Well not quite game over. We have Commodore to thank for taking onboard a dream machine that by rights should have sunk into obscurity. Jack Tramiel probably would still have used the 68000 if he had remained at Commodore


Jack Tramiel's Commodore was designing the 16-bit ZiLog Z8000-based C900. Key engineers from C900 moved to Tramiel's Atari Corp to lead the Atari ST project.

From C900 project, ex-Commodore employee Shiraz Shivji investigated different CPUs including the 32-bit National Semiconductor NS32000. Shiraz Shivji selected the 68000. ST TOS is largely DR's GEM GUI with GEMDOS 68K port.


@Matthey
Quote:

The 68000 dominated the 8086.

performance 68000
orthogonality 68000
GP registers 68000
address space 68000
datatypes 68000

code density tie (68k is better than x86 though)

support chips 8086
compatibility 8086
price 8086

Intel was lucky to land the IBM PC contract and they did so by selling at least early 8088 CPUs below cost so the high margin desktop market was far from instantaneous.

1. For lesser datatypes, 8086 wouldn't be wasting the clock access cycle like 68000's.


2. Intel offers 8087 FPU. An important aspect of the 8087 was the basis for the IEEE 754 floating-point standard.

68000 doesn't natively support floating point formats and must do it with software emulation.
68K's FPU arrives with 68881's 1984 release.

The 8087 provided two basic 32/64-bit floating-point data types and an additional extended 80-bit internal temporary format.


3. Intel fulfills IBM's 8-bit external bus with 16-bit ALU CPU 1980s era cost reduction requirement while 68008 is late.


4. IBM PC model 5150 supports 8087 FPU. https://21stdigitalhome.blogspot.com/2019/01/the-ibm-pc-5150-cpu-and-fpu.html

Compaq Deskpro PC clone had an 8086 processor.

Mac wouldn't support 68881 FPU until 1987's Macintosh II.
Amiga wouldn't support 68881 FPU until 1988's A2620 and A2000/020 models.
Atari wouldn't support 68881 FPU until 1990's Atari TT.

1984 released IBM PC/AT 286 supports 287 FPU option.
1986 released Compaq Deskpro 386 supports 387 FPU option.

Lotus 123 2.0's FPU support was part of the "killer app" for the PC platform.


https://imgur.com/a/coprozessoren-f-r-windows-3-1-9ocY2
The article says Win3.1 supports FPU and gives up to 80% more GUI performance. FPU usage for the GUI is also supported by MacOS.

For 387 FPU, WW0548.EXE update contains a revised Windows 3.1 WIN87EM.DLL file.
https://www.vogons.org/viewtopic.php?p=1063409#p1063409

Last edited by Hammer on 20-Dec-2024 at 12:00 PM.
Last edited by Hammer on 20-Dec-2024 at 11:58 AM.
Last edited by Hammer on 20-Dec-2024 at 11:52 AM.
Last edited by Hammer on 20-Dec-2024 at 11:50 AM.
Last edited by Hammer on 20-Dec-2024 at 11:47 AM.
Last edited by Hammer on 20-Dec-2024 at 11:31 AM.
Last edited by Hammer on 20-Dec-2024 at 11:29 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

 Status: Offline
Profile     Report this post  
kolla 
Re: what is wrong with 68k
Posted on 20-Dec-2024 13:38:32
#316 ]
Elite Member
Joined: 20-Aug-2003
Posts: 3357
From: Trondheim, Norway

@matthey

Quote:
I tend to believe that IBM and Commodore cross licensed Rexx and GUI IP. The history is being lost as the Amiga dies but there are non-Amiga sites saying the same thing.


Maybe you can find the two other pages too?

Please explain to me why such a license would exist? Amiga had ARexx by William S. Hawes, NOT IBM's REXX interpreter. And the look of OS/2 does not resemble anything Amiga at all. And which came first - OS/2 or ARexx?

_________________
B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC

 Status: Offline
Profile     Report this post  
Karlos 
Re: what is wrong with 68k
Posted on 20-Dec-2024 14:21:20
#317 ]
Elite Member
Joined: 24-Aug-2003
Posts: 4843
From: As-sassin-aaate! As-sassin-aaate! Ooh! We forgot the ammunition!

All this talk of ARexx reminds me, I always wanted to create a version of ARexx that had strong scalar type support and basic vectors and maps, while retaining the free form stem aggregate structures.

It would be nice if also compiles down to a more machine interpretable intermediate form for execution too.

But *most* importantly, I'd call it TRexx, lol

_________________
Doing stupid things for fun...

 Status: Offline
Profile     Report this post  
bhabbott 
Re: what is wrong with 68k
Posted on 20-Dec-2024 20:11:11
#318 ]
Cult Member
Joined: 6-Jun-2018
Posts: 509
From: Aotearoa

@Hammer

Quote:

Hammer wrote:

Are you arguing 24-bit display is limited to SGI? Hint: PC's SVGA cards can display 24-bit color (true color) on various models.

No, of course I'm not arguing that 24-bit display was limited to SGI. That would be silly.

But Ton Roosendaal stated that he got an SGI (presumably an IRIS Indigo) - NOT a PC - in 1991 because "The market demanded higher quality graphics, especially 24 bits color. If Amiga would have made that step sooner…".

Why an SGI and not a PC? Because a PC wasn't good enough for him either.

But what about his claim that Amiga hadn't 'made that step sooner'?

The first 24-bit display devices for the Amiga came out in 1988. These were NTSC only and so probably not suitable for what Ton wanted, but it shows that 24-bit was coming to the Amiga so it wasn't a 'dead end'. By 1991 there were many 24-bit cards available. Most were aimed at video production, but some could do higher resolutions too. For example the PP&S Rembrant did up to 1024Ă—1024 in 32-bit.

These developments were well publicized so Ton should have known about them, suggesting there was more to it than just having 24-bit. And there was,
Quote:
Another reason was platform instability. Both the hardware and the Amiga OS were notoriously unstable. You couldn’t rely on it to work for more than a couple of hours...

In the first year of our company we hardly had commercial jobs making 3D animations. Doing practical things with 3D was too slow anyway, we mostly did this for fun in our spare time. In 1990 we had our first 3D animated jobs, and it was clear we would never be able to offer the quality the industry would demand. So it wasn’t about looking at other Amiga software for inspiration, at that time I was already eyeing Alias, Wavefront, Softimage… the stuff used in Hollywood and by big TV stations...

Plenty of Amigas were used successfully by 'big TV stations', and not everyone had his experience of the Amiga being 'notoriously unstable'. But Hollywood was a different story. If they wanted production quality images they would need the best system they could get. In 1991 SGI released the IRIS Indigo 'budget' MIPS-based machine that was just the ticket, for ~ÂŁ9,000 (AU$18,000).

Like I said, this is so far outside the realistic expectations for a mass-market home computer that it's just silly - but typical of Amiga fans.

Quote:
Porter specifically wanted 1000 by 800 resolution with 8 bit planes and 16 million colors—something to exceed the current competition.

Of importance would be keeping the new video technology compatible with existing commercial Amiga software.

All of these plans would be discussed at Commodore’s worldwide engineering meeting, scheduled for September 22, 1987 at the Embassy Suites hotel in New York.

1. 1000 by 800 resolution, 8 bitplans and 16 million colors. AGA has 8 bitplanes and 18-bit HAM8 (lossy color compression) display capability.

2. Backward compatible.

3. Jeff Porter's advocacy before September 1987. Dale Luck advocated for a 16-bit chunky display, which is similar to 3DO's 16-bit chunky color sustained display performance mode.

Nothing wrong with floating ideas for future products. But in 1987 the A500 had just been released and was still to prove its worth - which it did, showing that the way forward was not trying to compete against high-end workstations.

The final realization of AGA as a low cost enhancement of the existing chipset was the right one. Just a pity they hadn't realized that earlier and scaled back their aspirations to an achievable goal which would fit the market. How nice would it have been to get AA in 1990 instead of the A3000, in a package similar to the A1200 and/or A600? It might even have made the CDTV worthwhile.

But no, they had to have their Unix workstation lookalike while they flailed around trying to make AAA work. And if you were there I bet you would have cheered it on too, because that's the way Amiga fans think.

 Status: Offline
Profile     Report this post  
bhabbott 
Re: what is wrong with 68k
Posted on 20-Dec-2024 20:56:53
#319 ]
Cult Member
Joined: 6-Jun-2018
Posts: 509
From: Aotearoa

Quote:

kolla wrote:
@matthey

Quote:
I tend to believe that IBM and Commodore cross licensed Rexx and GUI IP. The history is being lost as the Amiga dies but there are non-Amiga sites saying the same thing.


Maybe you can find the two other pages too?

Please explain to me why such a license would exist? Amiga had ARexx by William S. Hawes, NOT IBM's REXX interpreter. And the look of OS/2 does not resemble anything Amiga at all. And which came first - OS/2 or ARexx?

REXX was developed by IBM and open-sourced. AREXX was just one implementation, and there was zero reason to 'cross license' it back to IBM.

However IBM did have a reason to 'cross license' other Amiga IP, particularly patents. You never know when some obscure patent might come in handy, as we found out with the infamous XOR troll. According to Dave Haynie "We knew they wanted a license on the Amiga patents to keep them out of trouble if we came after a patent on them... It was just one of those general 'cover your ass and get some money out of everyone else in the business' deals".

OS/2 was designed to be a protected-mode successor of PC DOS for the 286. IBM intended to use it in their PS/2 computers, released in 1987. Its system calls were modeled after DOS and it had a text-based interface. Later on it got a GUI courtesy of Microsoft, similar to the GUI they used in Windows 1.0 in 1985.

So there was no connection between OS/2 and Amiga OS. Any similarity was purely coincidental - though both were (obviously) influenced by other stuff around at the time.

 Status: Offline
Profile     Report this post  
Hammer 
Re: what is wrong with 68k
Posted on 20-Dec-2024 22:58:20
#320 ]
Elite Member
Joined: 9-Mar-2003
Posts: 6171
From: Australia

@bhabbott

Quote:

No, of course I'm not arguing that 24-bit display was limited to SGI. That would be silly.

But Ton Roosendaal stated that he got an SGI (presumably an IRIS Indigo) - NOT a PC - in 1991 because "The market demanded higher quality graphics, especially 24 bits color. If Amiga would have made that step sooner…".

Why an SGI and not a PC? Because a PC wasn't good enough for him either.


The other part of Roosendaal's SGI selection is OS stability from memory-protected Unix-based IRX. This is for software development with 24-bit graphics.

In 1991, the Windows 3.1 NT release build was running late since it was in beta state e.g.
https://betawiki.net/wiki/Windows_NT_3.1_October_1991_build

Relative to SGI's MIPS R4000 platform, the PC was about 1.5 to 2 years late. This is why Intel demo'ed Pentium and 100Mhz 486 in 1992 and Microsoft keeps showing Windows NT 3.1 preview builds.

Microsoft has development control for Visual C++ and Windows NT at the same time.

The threats from the MIPS platforms are real.

A3000 had AMIX with 256 color A2410 and weak 68030-25 workstation CPU.
https://amiga.resource.cx/exp/a2410

On professional GUI apps, IRIS wasn't a barren wasteland like on AMIX.

X86 PC collective from Compaq, Intel, AMD to Microsoft have worked together to defeat MIPS, specifically SGI.

Quote:

@bhabbott

But what about his claim that Amiga hadn't 'made that step sooner'?

The first 24-bit display devices for the Amiga came out in 1988. These were NTSC only and so probably not suitable for what Ton wanted, but it shows that 24-bit was coming to the Amiga so it wasn't a 'dead end'. By 1991 there were many 24-bit cards available. Most were aimed at video production, but some could do higher resolutions too. For example the PP&S Rembrant did up to 1024Ă—1024 in 32-bit.

Amiga's non-RTG 24-bit cards are useless for Amiga's general-purpose application use cases.

Amiga's 1st RTG solution was GVP's EGS which was released in 1992. Commodore initiated FUD marketing against GVP's EGS with the promise of Commodore RTG solution while the real situation was Commodore was going bust with many engineers being fired during 1993.

On PC, the TIGA-based card would be useful for Windows apps and IBM 8514 DOS-compatible apps.

If Commodore was alive, it would have initiated FUD marketing against Phase 5's CGX RTG.

Many ex-Commodore engineers were hired by SGI.

Quote:

@bhabbott

Plenty of Amigas were used successfully by 'big TV stations', and not everyone had his experience of the Amiga being 'notoriously unstable'. But Hollywood was a different story. If they wanted production quality images they would need the best system they could get. In 1991 SGI released the IRIS Indigo 'budget' MIPS-based machine that was just the ticket, for ~ÂŁ9,000 (AU$18,000).

https://jmsnews.com/messages/message?id=4697
Babylon 5 switched to Windows NT-based PCs and Alphas in Season 2. Raytracing requires a lot of CPU power.

SGI was a victim of "death by a thousand cuts" from the commodity PC platform.

A cluster of cheap and mass-produced PCs as the poor man's render farm.

NVIDIA has driven the final nail on SGI with the start of GeForce 256's release. Economies of scale matter. It's like the raw economic monster USA beating Japan in WW2.

Raytracing is a bottomless pit when it comes to throwing compute power at it and that's with a modern GeForce RTX 4090.

The pattern, Amiga is used as a starter computer to earn some money before buying faster Pentium PCs or Alphas or MIPS or PowerMacs (from 1994).

Video Toaster's 24bit graphics weren't general purpose for Amiga apps when compared to RTG capable Macs (e.g. Quadra 840av) or Windows NT-based PCs.

Amiga's non-RTG 24-bit cards did nothing to improve the Amiga platform's 3D readiness.

Quote:

@bhabbott
Nothing wrong with floating ideas for future products. But in 1987 the A500 had just been released and was still to prove its worth - which it did, showing that the way forward was not trying to compete against high-end workstations.

Commodore didn't evolve the A500 product class fast enough and it was treated like a Coke product.. Henri Rubin didn't focus on Amiga's core graphics business.

https://www.youtube.com/watch?v=P4VBqTViEx4
Steve Jobs on the role of product and marketing people.

After A500's release, the rot killed the Amiga.
After 6502's release, the rot killed the 65xx CPU family from the microcomputer desktop market.

IBM's 386 delay was due to a marketing decision to maintain IBM's existing product stack i.e. 32-bit desktop conflicts with 32-bit minicomputers.

A600's execution was due to a marketing decision to create low Amigas with PCs for mid to high-end. Jeff Frank argued for high Amigas to be canceled. Blame Jeff Frank.

Ex-Amiga engineers led 3DO can display 24-bit graphics along with sustained performance mode 16-bit graphics. Ex-Amiga engineers led 3DO and also created 3DO M2.

Commodore should have kept the original Los Gatos Amiga team. The core reason why Amiga is great is the original Amiga engineers and Commodore's mass production engineers. The two halves balance each other.

Steve Jobs statement against "marketing people" is also applicable to Jack "marketing" Tramiel.

Apple CEOs after Steve Jobs left Apple couldn't replace MacOS until Steve Job's MacOS X. Good leadership matters. The problem with 68K was leadership at Motorola.

Last edited by Hammer on 20-Dec-2024 at 11:53 PM.
Last edited by Hammer on 20-Dec-2024 at 11:35 PM.
Last edited by Hammer on 20-Dec-2024 at 11:28 PM.
Last edited by Hammer on 20-Dec-2024 at 11:20 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

 Status: Offline
Profile     Report this post  
Goto page ( Previous Page 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 Next Page )

[ home ][ about us ][ privacy ] [ forums ][ classifieds ] [ links ][ news archive ] [ link to us ][ user account ]
Copyright (C) 2000 - 2019 Amigaworld.net.
Amigaworld.net was originally founded by David Doyle