@Hammer

Quote:

Hammer wrote: @cdimauro Quote: This IPC "measure" is only for single-byte instructions. All 65xx performance is dominated by the number of memory accesses. Which means that if you've an ADD instruction using the absolute addressing mode, it requires 3 bytes for its encoding -> 3 cycles for accessing those bytes. Plus one cycle for accessing the memory content. Total: 4 cycles (at least. Since crossing the page on the 6502/8502 added another cycle). In short: no, the IPC is absolutely not the one that you've reported. 65CE02 has 8-bit ALU, hence one shouldn't expect more than this.

Not really: it makes some 16-bit calculations (e.g.: addresses) without any penalty (whereas 6502/8502 incur into one additional cycle for cross-pages AKA more-than-8-bits-need-to-be-update).

However, yes: in general it's limited by its 8-bit ALU.
Quote:

65CE02 would be multi-cycling anything beyond 8 bits via the programmer's effort which is no different from multi-cycle 68000 implementation. The difference is that 68000's microcode is less verbose with memory bus when processing 16-bit and 32-bit datatypes which gives opportunities for Amiga's custom chipset value add.

The (great) difference is that the 68000 doesn't put on the programmers' shoulders the calculations which go beyond 8 bits.
Quote:

16-bit memory segmentation is wasteful beyond 65K memory.

For all 65xx processors.
Quote:

My point is Commodore LSI group CPU team has been degraded.

I'd make it shorter: Commodore LSI group team has been degraded.
Quote:

Between 65CE02 vs 68000, I prefer 68K for easier ASM due to the lease amount code for the job.

I've yet to find someone which thinks differently. Someone sane, I mean.
Quote:

68K was frozen in time which lacks 3D-related optimizations and focus, hence it's an administration issue with 68K's problems i.e. management problems.

The 68k can be easily extended in a natural, coherent way to add everything which is missing.
Quote:

Removing the microcode layer, a 68000 microarchitecture CPU core is a simpler design in its given era.

Yes, to some extent, for the 68000. No for the 68020+, because Motorola complicated too much the ISA.

Last edited by cdimauro on 28-Aug-2024 at 05:52 AM.

@thread

So a quick couple of Google searches reveal that no one ever released a computer using the Z80000.
Shame, we'll never find out how well it could've run Doom.

_________________
All the way, with 68k

@agami

No, but we could speculate the shit out of it ...

_________________
Doing stupid things for fun...

@agami

It's pointless when Z80000 has supply problems and it's late. Z8000/Z80000 didn't replicate X86's backward compatibility scale.

_________________
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 7900X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB

@agami

Actually, Olivetti did. S6000 later machines had Z80000 and L1M series 70+. These were large multiuser systems. But no manuals survive on the web. There were a couple of niche machines built by an autodiagnostic company that used Z80000 + Z8070(FPU), but none seem to have survived.

Last edited by WolfToTheMoon on 29-Aug-2024 at 09:49 AM.

_________________

@Hammer

Quote:

Hammer wrote: @agami It's pointless when Z80000 has supply problems and it's late.

Late for what, Doom?

While it may have been late to effectively compete with one CPU, it doesn't mean that it couldn't compete in other areas against another CPU. I was just a little surprised that it didn't find some sort of market, however niche it might have been.

_________________
All the way, with 68k

@WolfToTheMoon

Quote:

WolfToTheMoon wrote: @agami Actually, Olivetti did. S6000 later machines had Z80000 and L1M series 70+.

I'm surprised these didn't come up in my Google search results, especially because Google did return Z8000 results, and among them references to Olivetti machines using a Z8000.

I wonder if Sinclair/Amstrad was doing better at the time if they would've looked to build something on on the Z80000. I have to say that I'm kind of astonished that the 'Spectrum Loki' project was still based on a Z80. Aiming to build an "Amiga killer", I would've thought they'd at least use a Z8000 for that one.

_________________
All the way, with 68k

@agami

Quote:

agami wrote: @Hammer Quote: Hammer wrote: @agami It's pointless when Z80000 has supply problems and it's late. Late for what, Doom? While it may have been late to effectively compete with one CPU, it doesn't mean that it couldn't compete in other areas against another CPU. I was just a little surprised that it didn't find some sort of market, however niche it might have been.

Z80000 is too late for the Amiga platform.

PC's Doom success was built from 32-bit 386 and VGA install base foundation that started from 1987. Intel's revenues have shifted to a 486 majority from 1992!


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

AMD dominated 386DX sales with Am386-40.

From 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 Amiga didn't have "stable high-res" for business with economies of scale e.g. A2024's 5000 unit run. LOL

Last edited by Hammer on 30-Aug-2024 at 03:18 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 7900X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB

@agami

Quote:

agami wrote: @WolfToTheMoon Quote: WolfToTheMoon wrote: @agami Actually, Olivetti did. S6000 later machines had Z80000 and L1M series 70+. I'm surprised these didn't come up in my Google search results, especially because Google did return Z8000 results, and among them references to Olivetti machines using a Z8000. I wonder if Sinclair/Amstrad was doing better at the time if they would've looked to build something on on the Z80000. I have to say that I'm kind of astonished that the 'Spectrum Loki' project was still based on a Z80. Aiming to build an "Amiga killer", I would've thought they'd at least use a Z8000 for that one.

As for Spectrum Loki, the Z8000 is not backward compatible with the Z80, hence it killed any Z80 platform evolution.

Thanks to Zilog's incompatible 16-bit Z8000, Z80 CP/M reached a premature dead end.

16-bit Z280 was released in 1987 which is too late to revive the Z80-based platforms.

Z180 was released in 1997.

Idiots are running Zilog.

Last edited by Hammer on 30-Aug-2024 at 05:18 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 7900X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB

@agami

Quote:

agami wrote: The argument could be made (but not won) that Commodore could've asked the Hi-Toro team to re-engineer the Amiga computer around a different CPU. As I found out via a question on this very forum: any other contemporary CPU within the ballpark price would severely compromise the potency of the custom chipset. The more I've ruminated on the whole sordid affair, the more I'm convinced that outside of the immediate return of screwing over Atari, the medium term returns were based on a bet that all the tech in the Amiga would get cheap enough in the near future for Commodore to have their C64 2.0 moment. They kept remaking and losing that bet. A similar bet was made by NeXT several years later. Moore's Law is not linear. A linear equation between technological improvement and cost of production is extrapolated in more of a Monte Carlo average of highly variable data, driven more by market trends rather than what is possible in silicon. It is capitalism after all, and no one wants to halve the price of item if they can get away with it.

Bitplane-based graphics chipset with 16-bit (65KB) memory segmented model 65CE02 CPU is C65 from Commodore LSI group. 65CE02 is a little-endian.

The Amiga chipset is based on 68000's big endian flat memory model. For bare metal, another potential CPU family for the Amiga should be a big-endian capable.

Commodore LSI's C65 project started about 1985 as a response to Commodore-Amiga Inc's Amiga project.

From Commodore - The FInal Years,

Quote:

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. (skip) Commodore had purchased the rights from the Western Design Center (WDC) to produce Bill Mensch’s 65C02 for half price and had plans to use the chips in the LCD computer. However, there were restrictions. “We had the Western Design Center’s database for the 65SC02, but it had with it stipulations that it could not be sold as a stand-alone part,� says Gardei. If CSG wanted to sell the CMOS chips to Atari or other customers, it would have to design its own. (skip) Commodore’s Large Scale Integration department, headed by Bob Olah, began preliminary investigations in July 1985 into a CMOS 6502 chip which Lenthe called the 55C02. (He eventually shortened the name to 5502.) The LSI group would design the chip using 3 micron CMOS. (skip) Bill Gardei had worked for the LSI group since the beginning of the year. “I was interviewed and hired by Bob Olah in February of 1985,� he says. On his own, he decided to resurrect the 5502 project, this time using 2 micron CMOS. He dubbed the new chip the 4502. “All new CMOS custom chips were given the series number of 4000 and up, so 4502 was the obvious choice,� he says. “The 4502 core development started in late 1985. The original designers were myself and Charles Hauck.�

In 1985, Commodore had 3-micron which was followed by 2-micron fabs which are not exploited for the Amiga i.e. stuck in the 5-micron process node-based design.

There is no unified direction.

_________________
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 7900X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB

@agami

Quote:

I'm surprised these didn't come up in my Google search results, especially because Google did return Z8000 results, and among them references to Olivetti machines using a Z8000

Those very very niche machines. Apparently both Z8000 and Z80000 found a market in military applications - it seems some Z8000 MCUs are still used for engine management in jets.

Quote:

I wonder if Sinclair/Amstrad was doing better at the time if they would've looked to build something on on the Z80000. I have to say that I'm kind of astonished that the 'Spectrum Loki' project was still based on a Z80. Aiming to build an "Amiga killer", I would've thought they'd at least use a Z8000 for that one.

I could imagine a Z8000 derivative for the QL, as it was already gimped by the 8 bit bus and the Z8000 was cheaper and you have a cheap 32 bit option going forward. But hindsight is 20/20 and 68000 was a popular choice in the mid 80s, so you could expect more software to come by simply porting from other platforms.

Last edited by WolfToTheMoon on 30-Aug-2024 at 08:32 AM.
Last edited by WolfToTheMoon on 30-Aug-2024 at 08:30 AM.
Last edited by WolfToTheMoon on 30-Aug-2024 at 08:30 AM.
Last edited by WolfToTheMoon on 30-Aug-2024 at 08:29 AM.

_________________

@Hammer

Quote:

Hammer wrote: the Z8000 is not backward compatible with the Z80, hence it killed any Z80 platform evolution. Thanks to Zilog's incompatible 16-bit Z8000, Z80 CP/M reached a premature dead end.

That's not true. IBM killed CP/M when they got Microsoft to write PC DOS for them. Business users flocked to the PC because it was IBM, and they used PC/MS DOS because it was the PC's 'standard' OS.

Zilog didn't stop making the Z80 in 1979 when they introduced the Z8000. Think about how many millions of Z80 based machines were sold since then, then think about how many of them ran CP/M. The market for Z80s was huge, but the market for CP/M business computers was small and shrunk after the PC arrived. That didn't stop Amstrad from selling millions of CPCs and PCWs with CP/M, but by 1984 most businesses were using PCs.

If the Z8000 had been 100% Z80 (or 8080) compatible it wouldn't have made any difference. How do we know? Intel made a big deal about how 8080 asm source could be transcoded into 8086 asm, but few if any developers bothered to do it. Applications in those days were fairly easy to simply rewrite from scratch, which was a good idea because the PC had numerous features that CP/M machines didn't. Few customers would want to run clunky CP/M apps when the new IBM PC equivalents were so much better.

In 1984 NEC introduced the V20, which as well being 8086 compatible could also execute 8080 code. But did anyone use that feature? In a market where PC DOS applications were exploding and CP/M was dying It was pointless.

The Z8000 'failed' for several reasons. IBM chose the 8088 instead because it was cheaper and backed by Intel, and it arrived only a few months before the more powerful 68000 which was its other competitor. Several Unix systems used the Z8000. The 64k variant was used in military applications up until 2012. But these were low volume markets. The only 'mass-market' desktop computer to use it was the Olivetti M20, of which ~50,000 units were sold in the first year of production. But the M20 didn't become popular for the usual reason - it wasn't IBM compatible.

Quote:

16-bit Z280 was released in 1987 which is too late to revive the Z80-based platforms.

Z80-based systems were showing their age like most 8-bit systems, but in 1987 those platforms didn't need 'reviving' because were still going strong. Amstrad introduced the CPC6128 in 1985 and the PCW9512 in 1987. In 1986 they acquired Sinclair and introduced the ZX Spectrum +2, followed by the +3 in 1987. MSX2 was introduced in 1985. The Sega Master System was launched in the US in 1986 and Europe in 1987. ~10-13 million were sold. The Sega Genesis / Mega Drive also had a Z80 in it. Launched in the US in 1989, it sold over 30 million units.

Quote:

Idiots are running Zilog.

The people at Zilog weren't idiots, they just didn't have the resources of Intel or Motorola. In any market where compatibility is a major issue, generally only one or two competing 'standards' survive. Number 3 doesn't become come as popular and so dies. Just because Zilog's offering didn't 'hit the spot' like the 8088 and 68000, doesn't mean the company was run by idiots. They didn't quite manage to beat Intel and Motorola in the 16/32 bit CPU game, but the Z80 was so successful that Zilog only retired it this year, after 48 years of production.

@Hammer

Quote:

Hammer wrote: @agami ... the Z8000 is not backward compatible with the Z80 ...

I know. Neither was the 68000 backwards compatible with the 6800.

Many of the 16/32-bit microprocessor designs of the late '70s and early '80s were not compatible with their 8-bit predecessors. It's what tends to happen in the early phase of any new domain: There's always a bit of "Wild West" until some standards emerge.

Same as Macintosh to Apple II, and Atari ST to Atari 400/800, the Amiga was not backward compatible with the C64 or the C128, therefore it would kind of make sense that any "Amiga-killer" computer needn't be backward compatible with the paltry Spectrum 48. I don't think that backward compatibility with an 8-bit system was a killer feature in the second half of the '80s.

_________________
All the way, with 68k

@agami

Quote:

 I don't think that backward compatibility with an 8-bit system was a killer feature in the second half of the '80s.

I think this is where you're wrong. People who bought these cheap home computers would find it very helpful if their existing software run on their new computer. It would have been an competitive advantage against most other 16/32 bit home computers which had no software at launch.

Let's look at C= range in 1985
C64 sold at 150$
C128 debuted at around 300$
Amiga 1000 debuted at around 1500$

For comparison, the ST was around 600$ at launch.

Let's make a parallel universe in which C128 is not made, but rather a 16 or 32 bit 6502 compatible which is priced somewhere in between C128 and ST, but probably closer to historical C128 prices. Now you have a system which is capable of future growth, cheaper than all of the new competition and, unlike all other new 16/32 bit home computers(not counting the PC), already has a large software library and owners of C64 could reuse their software.
Tell me that a computer like that doesn't make more sense than A1000 did.

Last edited by WolfToTheMoon on 31-Aug-2024 at 08:56 AM.

_________________

@WolfToTheMoon: if this computer was only about about an upgraded CPU, then it doesn't make more sense.

The Amiga 1000 was NOT only about the processor. Rather, the chipset was its major advantage compared everything.

This great computer defined the "multimedia" concept as we know today, as well as what's a personal computer with a modern OS.

In short: the Amiga 1000 was a computer of another class.

@cdimauro

Quote:

@WolfToTheMoon: if this computer was only about about an upgraded CPU, then it doesn't make more sense.

who says it should be only about the upgraded CPU?

Quote:

In short: the Amiga 1000 was a computer of another class.

that is true, but it was also of another class in price.

_________________

Hammer Quote:

In 1985, Commodore had 3-micron which was followed by 2-micron fabs which are not exploited for the Amiga i.e. stuck in the 5-micron process node-based design. There is no unified direction.

Amiga Corporation had already chosen the process and performed the chip layout using a very conservative and economical 5000nm NMOS process from the mid 1970s. There may have been auto software layout tools by that time but they were likely expensive and limited. Moving to a new process was not as simple as selecting the chip process and auto layout in software as is possible for large fully synthesizable logic chips today. It made sense to produce the first Amiga chipset using the antiquated chip process that it was designed for. The Amiga chipset was still a game changer due to integration and low cost despite the antiquated chip technology used. It made sense to upgrade the Amiga chipset to a more modern CMOS process when enhancing the chipset, as well as further integrating the chipset into fewer chips. Agnus and Denise likely could have been a single CMOS chip years before ECS was released but CBM management may have wanted drop in chip upgrades. This doesn't explain why Alice and Lisa were not integrated for AGA but we know AGA was a rush job which likely explains it. In other words, negligent CBM planning by management.

bhabbott Quote:

The Z8000 'failed' for several reasons. IBM chose the 8088 instead because it was cheaper and backed by Intel, and it arrived only a few months before the more powerful 68000 which was its other competitor. Several Unix systems used the Z8000. The 64k variant was used in military applications up until 2012. But these were low volume markets. The only 'mass-market' desktop computer to use it was the Olivetti M20, of which ~50,000 units were sold in the first year of production. But the M20 didn't become popular for the usual reason - it wasn't IBM compatible.

The Z8000 was likely cost competitive with the 8088.

year | CPU | transistors | pins
1975 6502 3,510 40
1976 Z80 8,500 40
1978 8086 29,000 40
1979 8088 29,000 40
1979 Z8001 17.500 48
1979 Z8002 17,500 40
1979 68000 64,000 64
1982 68008 70,000 48

The Z8002 has the cost advantage over the 8088 but IBM wanted addressing beyond 64kiB which required the Z8001 with more pins partially offsetting the transistor advantage. The 8088 is a gimped 8086 like the 68008 is a gimped 68000 that has minimal cost advantages. Intel just made IBM a good deal on a gimped loss leader 8088, likely hoping that IBM would need to upgrade such a gimp CPU. IBM did need to upgrade their gimp 8088 CPU, forever changing history.

bhabbott Quote:

The people at Zilog weren't idiots, they just didn't have the resources of Intel or Motorola. In any market where compatibility is a major issue, generally only one or two competing 'standards' survive. Number 3 doesn't become come as popular and so dies. Just because Zilog's offering didn't 'hit the spot' like the 8088 and 68000, doesn't mean the company was run by idiots. They didn't quite manage to beat Intel and Motorola in the 16/32 bit CPU game, but the Z80 was so successful that Zilog only retired it this year, after 48 years of production.

The Z80 ate Intel's lunch early like the 68000 did later.

Intel vs Z80 catch up: 8080 to 8085 to incompatible 8086
Intel vs 68000 catch up: 8086 to 80186 to 80286 to 80386

Nothing a little Operation Crush propaganda and luck can't fix though.

The Z80 finally being out of production is only the original Z80. Embedded variants like the eZ80 are still used in MCUs. A 6502 family MCU may be used where there is minimal code and ultra low power requirements but the Z80 has a much better ISA including better orthogonality, more registers, more powerful addressing modes and better code density than the 6502 or 8080/8085. It was introduced only a year after the 6502 too.

@WolfToTheMoon

Quote:

WolfToTheMoon wrote: @cdimauro Quote: @WolfToTheMoon: if this computer was only about about an upgraded CPU, then it doesn't make more sense. who says it should be only about the upgraded CPU?

You were only talking about the CPU.

Anyway, who should have upgraded the C64 chipset? The same which created the 128? The Plus/4?

The engineers which created the VIC-II and SID left the company, and we've seen what kind of "beautiful" things made the ones which stayed or the new hires.

Commodore has bought the Lorraine team exactly because they were completely missing such expertise. With the notable difference that Lorraine = Amiga went well beyond the expertise of the time, giving to the history a clear revolution.
Quote:

Quote: In short: the Amiga 1000 was a computer of another class. that is true, but it was also of another class in price.

Then any comparison doesn't make sense, right?

WolfToTheMoon Quote:

who says it should be only about the upgraded CPU?

The 68k Amiga had the evolutionary CPU and chipset. The problem was the CBM dinosaur that became extinct with the technology.

WolfToTheMoon Quote:

that is true, but it was also of another class in price.

The CBM 6502 evolution did not begin with the C64 but the Pet.

year | PC | price | price with inflation
1977 Pet $795 ~$4,126 (1st gen evolution)
1982 C64 $595 ~$1,939 (1st gen cost reduction)
1985 A1000 $1285 ~$3,756 (2nd gen evolution)
1987 A500 $699 ~$1,935 (2nd gen cost reduction)

The inflation adjusted prices are very close for these legendary evolutionary 6502 and 68k PCs. The more advanced 68k Amiga design allowed high compatibility to be retained for not just the cost reduced followup but for far more advanced designs that may be nearly unrecognizable compared to modern hardware if using very cheap to mass produce modern silicon.

Last edited by matthey on 31-Aug-2024 at 06:18 PM.
Last edited by matthey on 31-Aug-2024 at 06:18 PM.

@cdimauro

Quote:

You were only talking about the CPU. Anyway, who should have upgraded the C64 chipset? The same which created the 128? The Plus/4?

Why not the C128 team? Graphics and sound of the C128 were not competitive with the chipset of the A1000, but they were not 5 times worse - and C128 was 5 times less expensive(and only had half the RAM of the A1000).

A 16/32 bit C64 compatible home computer would have sold extremely well in 1985 and later... there were already millions of C64s out there and a lot of software. It was a good base to build upon. On the other hand, while the A1000 was amazing for it's capabilities at the time, it had no software - simple as that. It took another 3-4 years for the Amiga platform to take off. In those 3-4 years, you can sell a lot of C64 16/32 bit compatible computers - I mean, even the C128 sold a few million units, only outsold by the C64 and A500.

_________________