Multiple selection allowed. PPC survived longer than the other big fat dinosaurs but is it extinct?

The order below is approximate code density from worst at the top to best at the bottom.

architecture | type | status | markets
IA-64 VLIW *dead*
Alpha RISC *dead*
PA-RISC RISC *dead*
MIPS RISC *dead*
m88k RISC *dead*
SPARC RISC *dead*
PPC RISC *dead*
ARM RISC deprecated embedded
6502 accumulator alive embedded
ARM64/AArch64 RISC alive embedded
x86-64 CISC alive desktop|server
RV64IMC RISC alive embedded
SuperH RISC revived(J Core) embedded
x86 CISC deprecated desktop|server
RV32IMC RISC alive embedded
Thumb RISC alive embedded
Thumb-2 RISC alive embedded
Z80 accumulator alive embedded
m68k CISC neglected retro

http://deater.net/weave/vmwprod/asm/ll/ll.html

Last edited by matthey on 04-Nov-2024 at 09:40 AM.

Poll : has this survey (or even Matthew as usual) got anything interesting to say?

_________________
PowerMac G5 2,7Ghz - 2GB - Radeon 9650 - MorphOS 3.14
AmigaONE X1000, 2GB, Sapphire Radeon HD 7700
FPGA Replay + DB 68060 at 85Mhz

@matthey

Higley biased.

IBM uses Power CPU’s

PowerPC is basically half a Power chip, so no Power is not dead, there for PowerPC is not, it might not be popular CPU architecture but that is not the point.

The choice of CPU for AmigaOS uses has been based on price per chip, this is why we end up with chips that is EOL, because that’s what we get on a discount, and this why we ended up with the PA-SEMI, QorIQ P1022, and AMCC chips. These are chips generations behind, the best chips.

It does not help that development time takes far too long in Amiga land, but the thing is AmigaONE computers are normally fully usable Computers, VOX might disagree, but Kas1e did buy VOX computer and did get it working, not everyone is happy with what they got, or it’s not always a happy story when things are broken. And I agree that it is shameful how some have been treated by the dealers. In the end I think it hurts the dealerships.

The choice of chips feels like self-sabotage, at least in the case of P1022, that ended in huge amount of wasted time on FPU emulation and workarounds. And I feel that was also the case with Eye-Tech computers, their big performance difference between Pegasus II and AmigaONE-XE, so at the end of the day I can’t say working around the problem with software was not the right choice, again a big waste of time. This is my main concern, we have limited numbers of developers, and we can’t afford to waste their time. by making bad choices. In particular when they are not working full time.

Last edited by NutsAboutAmiga on 03-Nov-2024 at 08:05 PM.
Last edited by NutsAboutAmiga on 03-Nov-2024 at 08:02 PM.
Last edited by NutsAboutAmiga on 03-Nov-2024 at 08:00 PM.

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

@matthey

Your questions are not really useful because:
The term 'PPC' has been abandoned, even by IBM.

If you are talking of PPC,
- do you mean 'Power as Architecture' ?
- do you mean Power hardware as Amiga successor ?
- do you mean Power OSes like MOS, AROS, Linux with UAE as Amiga successor ?
- do you mean applications for AOS4, MOS or other Power based enhancement ?

One thing I can say for sure:
PPC, Power, 68k, Coldfire have been abandoned by Motorola aka Freescale aka NXP ...

Last edited by OneTimer1 on 03-Nov-2024 at 08:01 PM.
Last edited by OneTimer1 on 03-Nov-2024 at 08:00 PM.
Last edited by OneTimer1 on 03-Nov-2024 at 07:58 PM.

@all
The poll was a joke based on a post about PPC by ppcamiga1. PPC was a subset of POWER but a distinct architecture in the same way that ColdFire was a subset of the 68k but not completely compatible. PPC is more dead than the 68k with most, and likely all, of the development coming for retro cores. I believe there are more retro 68k cores in development than PPC cores, partially because PPC cores capable of simulating PPC console CPUs require a large expensive FPGA. There are a few 68k cores in FPGA that have tried to push beyond the performance of the newest 68060 but the 68060 silicon is 30 years old. The newest PPC silicon is about 10 years old and Amiga1 PPC CPUs are older. PPC silicon is older than 68k silicon was when it was replaced. The 68k was still the most popular 32-bit embedded CPU by volume up to about 2000 so it was far from dead when it was first being replaced by PPC and less dead than PPC is today after quickly being replaced by the similar but improved AArch64.

RISC Volume Gains But 68K Still Reigns
https://websrv.cecs.uci.edu/~papers/mpr/MPR/19980126/120102.pdf

POWER isn't really an option for the Amiga by the way. IBM tried to make POWER9 more modular and cheaper so it would scale down to workstation levels but they abandoned it because big business customers wanted beefier CPUs and more processing power. POWER lives for now because of backward compatibility despite large SMT cores being expensive to develop and produce. It's a really bad choice for an AmigaOS that doesn't even have SMP support after more than a decade.

@matthey

There's no positive options towards PPC (which will antagonise some people) so this isn't anything other than flamewar bait IMHO! You could in theory argue that although there is no development of PPC cores anymore the A1222+ and X5000/40 are for sale still so I guess that's a potential poll option just to be fair?

Last edited by BigD on 03-Nov-2024 at 09:59 PM.

_________________
"Art challenges technology. Technology inspires the art."
John Lasseter, Co-Founder of Pixar Animation Studios

@matthey

Quote:

matthey wrote: RISC Volume Gains But 68K Still Reigns https://websrv.cecs.uci.edu/~papers/mpr/MPR/19980126/120102.pdf

Interesting document, that was back in 1998(?) 68k including Coldfire where widely used in the automotive industry, MIPS was bigger than ARM maybe because the Playstation and some Phones/PDAs used it.


Today MIPS successors are known under the name RISC-V.

@OneTimer1

Really crazy Motorola dropped the ball by favouring PPC over 68k for the desktop!

_________________
"Art challenges technology. Technology inspires the art."
John Lasseter, Co-Founder of Pixar Animation Studios

@thread

To those clutching at the "IBM still develops Power ISA CPUs" straw, I say you need a serious reality check.

From 2011 to 2013 (inclusive), Aston Martin produced and marketed a consumer market car, the Cygnet.
The Cygnet is long dead, and anyone holding onto the notion that it is not dead because Aston Martin still produces cars, is in dire need of the same reality check as people thinking that IBM's Power 10 or Power 11 cores will someday somehow filter down to compete with ARM and even RISC-V cores.

Or maybe you're still having that wet dream that someday someone will create the ultra-mega Amiga OpenPower workstation with 16-core/128-thread Power10 CPU, 4TB DDR5 RAM, OpenCAPI + PCIe 5.0 multi-GPU envy-making machine, so that you can ask intel if they like apples? Then it's skip Go and strait to therapy for you.

People still ride horses and ride in horse-drawn carts, but for the intents and purposes of being a common and practical contemporary mode of transport, horses are dead.

_________________
All the way, with 68k

@matthey

Your 6502 being alive embedded while ARM 32bit being deprecated embedded is a joke.

ARM is larger than the Western Design Center.

Just as X86-64 v1 (AMD64)'s patent expired in 2023, patent expiration also applies to ARM.

For ARM 64bit ISA, ARMv8's Y2011 is about 8 years behind AMD64.

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

Windows 11 24H2 has moved to X86-64 v2 hard minimum requirements i.e. Intel Nehalem (2008)/AMD Bulldozer/VIA Eden C level.

Most current game generation for desktop game consoles and PCs is at the X86-64 v3 level.

Last edited by Hammer on 03-Nov-2024 at 11:39 PM.
Last edited by Hammer on 03-Nov-2024 at 11:29 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 7900X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB

Embedded you say? A new computer was recently released...
https://www.commanderx16.com/

long live the 65C02!

But yes, since the 6502 can run on 20mA, it lives on happily in many heart pace-makers...
It also controls many a hot water heater in your basement...

@matthey

_________________
Indigo 3D Lounge, my second home.
The Illusion of Choice | Am*ga

@NutsAboutAmiga

VOX Agrees x1000 is fully usable computer UNDER LINUX.
Check facts, kas1e wanted is trophy, last NG he did not have, and I am glad I have managed
to sell to him in midst of new madness.

VOX sais please make OS4 fully use IBM POWER CPU and TALOS boards :DDDD

VOX says it would be great if it was OS4 dream machjine, but it isnt, {PPC isnt dead yet, but instead of using PPC zombiness to justify whatever, there is still path for OS3.x, AROS,MorphOS to go. Best in a concentrated effort. All OS4 backports welclomed.

Use PPC cause the Lord seh so
https://youtu.be/trZ0x6oXvmM

Last edited by vox on 04-Nov-2024 at 07:21 AM.
Last edited by vox on 04-Nov-2024 at 07:18 AM.
Last edited by vox on 04-Nov-2024 at 07:16 AM.

_________________
Future Acube and MOS supporter, fi di good, nothing fi di unprofessionals. Learn it harder way!

@agami

Quote:

agami wrote: From 2011 to 2013 (inclusive), Aston Martin produced and marketed a consumer market car, the Cygnet.

"marketed" is a stretch, "produced" an outright lie.

They slapped their badge on a Toyota in an effort to circumvent some regulations and not having to pay fines associated with it.

All these discussions are pointless as PPC, 68k, 16/32Bit x86, 6502, Z80 etc are all dead just at different stages of decomposition.

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

OneTimer1 Quote:

Interesting document, that was back in 1998(?) 68k including Coldfire where widely used in the automotive industry, MIPS was bigger than ARM maybe because the Playstation and some Phones/PDAs used it.

The article is from 1998 but the data is from 1997. Each PS1 had 1 MIPS CPU, each N64 had 2 MIPS CPUs, each Saturn had 3 SuperH CPUs and the 68k CPU volume easily outpaced the number 2 by volume MIPS and number 3 by volume SuperH combined. MIPS was gaining quickly on the 68k but it projected for MIPS to take years to catch the 68k.

RISC Volume Gains But 68K Still Reigns
https://websrv.cecs.uci.edu/~papers/mpr/MPR/19980126/120102.pdf Quote:

MIPS had an undeniably good year, more than doubling over the previous year yet again. At this rate, MIPS could topple the 68K from the volume throne in just two to three years. Maintaining that rate, though, may be nearly impossible. Much of the architecture’s success is tied to two systems that will soon peak in sales. Even though MIPS is broadening its base into new markets, none seems likely to repeat the successes of the PlayStation and N64. Handheld PCs, printers, and WebTVs just don’t have the volume potential the game systems had. Digital cameras may be the closest thing to a hit consumer product in 1998, but MIPS is ill represented in that application. Like the i960, MIPS may follow its top-RISC title with a long plateau.

The PS2 released in 2000 did surpass the PS1 in sales but that means the 68k likely remained the volume leader up to about 2000 which I stated in my first post of this thread (I'm not so sure MIPS dethroned the 68k although SuperH did before ARM many years later). Some Amiga businesses rushed to follow the Mac and replace the 68k with PPC but 68k volumes were about the size of the x86 PC market and PPC Mac desktop and embedded CPU volume combined look like they were less than 10% of the 68k volume. Motorola was pushing PPC hard and focusing on new PPC cores while only the ColdFire where PPC was too fat to scale down received much development. ColdFire was better liked than PPC for embedded use but developers were often frustrated by the unnecessary lack of 68k compatibility and unnecessary deep castration of the 68k ISA which was shown by later ColdFire ISAs bringing back more of the 68k ISA because it was useful after all. PPC eventually had embedded niches in high end auto and telecommunication hardware because ARM was so weak but Motorola had high performance 68k cores like the 68060 which could reach similar performance with a smaller footprint. Stopping 68k development while it was so popular was like CBM killing the Amiga 500 while it was still popular. Motorola/Freescale gave up the embedded market as a result even though it took many years for the 68k popularity to drop down to PPC levels.

OneTimer1 Quote:

Today MIPS successors are known under the name RISC-V.

Right. MIPS is officially dead too.

https://en.wikipedia.org/wiki/MIPS_Technologies Quote:

MIPS was founded in 1984 to commercialize the work being carried out at Stanford University on the MIPS architecture, a pioneering RISC design. The company generated intense interest in the late 1980s, seeing design wins with Digital Equipment Corporation (DEC) and Silicon Graphics (SGI), among others. By the early 1990s the market was crowded with new RISC designs and further design wins were limited. The company was purchased by SGI in 1992, by that time its only major customer, and won several new designs in the game console space. In 1998, SGI announced they would be transitioning off MIPS and spun off the company. After several years operating as an independent design house, in 2013 the company was purchased by Imagination Technologies, best known for their PowerVR graphics processor family. They were sold to Tallwood Venture Capital in 2017 and then purchased soon after by Wave Computing in 2018. Wave declared bankruptcy in 2020, emerging in 2021 as MIPS and announcing that the MIPS architecture was being abandoned in favor of RISC-V designs.

There are more politics and propaganda in architecture choices than technical evaluations.

Hammer Quote:

Your 6502 being alive embedded while ARM 32bit being deprecated embedded is a joke. ARM is larger than the Western Design Center.

Lou is right on this one. The 6502 will still be alive in 2996.



https://spectrum.ieee.org/the-truth-about-benders-brain

The 6502 is about as simple as a CPU gets which means extremely low power for battery use. The Z80 is a small step up in complexity and still alive in MCU form because it has a better ISA and better code density which may be better with more code. These CPUs are not disappearing anytime soon although they may be hidden in deeply embedded hardware because they are tiny cores.

The original ARM ISA is on the way out. ARM is starting to develop AArch64 only cores except for their Cortex-M cores.

https://en.wikipedia.org/wiki/ARM_Cortex-M#Deprecations Quote:

The ARM architecture for ARM Cortex-M series removed some features from older legacy cores: o The 32-bit ARM instruction set is not included in Cortex-M cores. o Endianness is chosen at silicon implementation in Cortex-M cores. Legacy cores allowed "on-the-fly" changing of the data endian mode. o Co-processors were not supported on Cortex-M cores, until the silicon option was reintroduced in "ARMv8-M Mainline" for ARM Cortex-M33/M35P cores. The capabilities of the 32-bit ARM instruction set is duplicated in many ways by the Thumb-1 and Thumb-2 instruction sets, but some ARM features don't have a similar feature: o The SWP and SWPB (swap) ARM instructions don't have a similar feature in Cortex-M. The 16-bit Thumb-1 instruction set has evolved over time since it was first released in the legacy ARM7T cores with the ARMv4T architecture. New Thumb-1 instructions were added as each legacy ARMv5 / ARMv6 / ARMv6T2 architectures were released. Some 16-bit Thumb-1 instructions were removed from the Cortex-M cores: o The "BLX " instruction doesn't exist because it was used to switch from Thumb-1 to ARM instruction set. The "BLX " instruction is still available in the Cortex-M. o SETEND doesn't exist because on-the-fly switching of data endian mode is no longer supported. o Co-processor instructions were not supported on Cortex-M cores, until the silicon option was reintroduced in "ARMv8-M Mainline" for ARM Cortex-M33/M35P cores. o The SWI instruction was renamed to SVC, though the instruction binary coding is the same. However, the SVC handler code is different from the SWI handler code, because of changes to the exception models.

Thumb and Thumb-2 live on for small ARM cores because AArch64 doesn't have good enough code density to scale that low. AArch64 has better code density than the original ARM ISA so it is dead just like PPC. Code density is important! I didn't post the list of architectures by code density above with all the dead ISAs with poor code density at the top for nothing. Most early RISC ISA developers just didn't realize how important code density was or foresaw caches outgrowing the logic of the cores. Hitachi got it right and was ahead of its time with the RISC SuperH ISA which resembles the CISC 68000 ISA they were producing and licensed to ARM resulting in the Thumb ISAs. The 68k created the 32-bit embedded market and ruled it for about two decades before neglect allowed SuperH to take over followed by ARM Thumb/Thumb-2. The successor embedded leaders were good code density ISAs with heritage going back to the 68k. The 68k still has the best performance metrics, including code density, of these compact ISAs.

Last edited by matthey on 04-Nov-2024 at 09:31 AM.
Last edited by matthey on 04-Nov-2024 at 09:24 AM.

@Lou

Quote:

The Commander X16 is The 8-Bit Guy’s dream computer, designed to evoke the same fondness and nostalgia many of us had for 8-Bit computers, while retaining closeness to the hardware from a programming perspective... CPU WDC 65C02S @ 8 MHz Audio Yamaha YM2151 sound chip (8 channels of 4-Op FM) PCM audio on VERA (Up to 48kHz, 16-bit stereo) PSG of 16 channels with 4 selectable waveforms “VERA� module specifications Video generator featuring: Multiple output formats (VGA, NTSC Composite, NTSC S-Video, RGB video) at a fixed resolution of 640x480 @ 60Hz Support for 2 layers, both supporting: 1/2/4/8 bpp tile and bitmap modes Support for up to 128 sprites (with inter-sprite collision detection). Will it be compatible with Commodore 64 software? No

Fail. The only thing 'nostalgic' about the hardware in this machine is the CPU, and even that isn't a period 6502. Not C64 compatible, only the developer version supports a 1541 drive, and it doesn't have a tape interface. So what use is it? I can't imagine anyone wasting their time developing software for it.

But back on topic, yes - as far as desktop computing is concerned PPC is dead. And now that AmigaOS 3.x development has restarted and solutions like PiStorm exist, PPC is practically dead on the Amiga too. That doesn't mean it shouldn't be part of the retro scene though, just that it shouldn't be promoted as the future of the Amiga. It was always a kludge that many of us avoided, and now we don't need it.

Time to declare that he's Dead Jim, and move on. We knew the redshirt would die, but Captain 68Kirk and the others will be with us in the next episode of Amiga Trek, boldly going where no AmigaOne has gone before.

@matthey

the survey looks to me very amiga-centric. PPC as desktop system is dead latest since Apple dropped it and changed platform. And in amiga context we talk about desktop hardware. If you look general then there is PPC still used in embedded solutions and in IBM servers. So the answer to the question depends on which market you look. It is retreating also in embedded market where ARM processor become more and more common but it still exists.

@matthey

Quote:

Some Amiga businesses rushed to follow the Mac and replace the 68k with PPC but 68k volumes were about the size of the x86 PC market and PPC Mac desktop and embedded CPU volume combined look like they were less than 10% of the 68k volume. Motorola was pushing PPC hard and focusing on new PPC cores while only the ColdFire where PPC was too fat to scale down received much development. ColdFire was better liked than PPC for embedded use but developers were often frustrated by the unnecessary lack of 68k compatibility and unnecessary deep castration of the 68k ISA which was shown by later ColdFire ISAs bringing back more of the 68k ISA because it was useful after all. PPC eventually had embedded niches in high end auto and telecommunication hardware because ARM was so weak but Motorola had high performance 68k cores like the 68060 which could reach similar performance with a smaller footprint. Stopping 68k development while it was so popular was like CBM killing the Amiga 500 while it was still popular. Motorola/Freescale gave up the embedded market as a result even though it took many years for the 68k popularity to drop down to PPC levels.

Are we able to learn form these history lessons? In the UK businesses have been dealt a blow by Employers National Insurance rates going up, minimum wage increase and further workers rights bureaucracy. Reaction Engines, a leading light of aeronautical and F1 engine design has gone into administration! It's like we can only perpetuate big stable dinosaurs like IBM and Intel and struggle to see good products from smaller companies and to invest in them! The only main ongoing influence the Amiga had in the mainstream was on DirectX development for Windows, the legacy of UK games industry (which Sony SIE - Corporate American PlayStation, is now set on destroying) and some early 3D rendering software and the later iterations like LightWave!

It's sad it's legacy wasn't bigger but at least it started with the 68k and faded before people remembered it as a sad PPC Mac wannabe. The AmigaOne project technically peaked with the X1000 which used PA-Semi chips that had already been snapped up by Apple! There was no second coming through PPC and only through ARM and new 68k FPGA designs like the 68080 Apollo core have we forged ahead with anything of lasting significance.

To demonstrate, I play the PS1 version of WipEout 2097 on THEA500 Mini using the Pandory500 softmod and not the Amiga PPC version on an overpriced AmigaOne or CyberstormPPC/BlizzardPPC board! Why would anyone do that! MorphOS continues to be a shining beacon of sensible compromise and may survive on Intel chips but the PPC period was pretty much a disaster for anyone other than Apple!

Last edited by BigD on 04-Nov-2024 at 09:57 AM.

_________________
"Art challenges technology. Technology inspires the art."
John Lasseter, Co-Founder of Pixar Animation Studios

@matthey

Not quite dead, POWER9 is still competitive with current Intel. But the AmigaOne systems of the past 20+ years have pretty much ruined the reputation of PPC as a popular CPU product. Just look at the A1222+: a SoC with no AltiVec or even MMU - makes PPC look like a joke to anyone unfortunate enough to stumble across it, and the other AmigaOne systems aren't really any better better.

So not extinct, I'm sure PPC is still competitive in some niche products, but the most likely consumer path into a PPC system, AmigaOne, is littered with low-performance systems nobody outside of a handful of Hyperion fanatics wants.

It goes without saying this isn't a great look for the Amiga world in general either.

@AmiRich

Quote:

AmiRich wrote: Not quite dead, POWER9 is still competitive with current Intel

If POWER9 systems where available for the price of a Wintel system, no one here would call PPC/Power a fail.
And if you where able to run AOS4, MOS and Linux on a VR (without CPU Emulation) most people here would be glad.