@OneTimer1

Quote:

OneTimer1 wrote: Quote: Kronos wrote: Also, just looked it up and it seems dual core Xeons were only released in 2004 so any switch before would have been a side step at best. Apple went from a 64 Bit platform to a 32 Bit platform ... Quote: Kronos wrote: Apple switched to PPC because 68k was deemed EOL by Motorola, something that did not happen to PPC before the Intel switch. Well the 68040 was not really that jump foreword that was needed by Apple, and Motorola didn't had any good ideas for faster chips, so a switch supported by Motorola was a logic consequence, like it seemed to be for Amiga. But when Motorola failed with a successor to G4 and a performant power saving variant for a Laptop, they had to look for something else. PPC might not have been EOL then but it didn't look so promising any more.

Exactly!!!

@cdimauro

Whatifs are pointless and trying to argue why something that did not happen would have happened isn't much different. But....

Quote:

cdimauro wrote: That was different. NextSTEP was already running on different architectures, and even on 68ks.

All but the x86 versions had been discontinued years before being bought by Apple.

- 68k -> PPC "meh" on native code, bad on legacy code -> MS had to bailout Apple a few years later

- PPC -> Intel (in 2005/6) "meh" for the lowend ( Core Solo), "big bang" for everything above, bad for GPU -> Mac gaming died and has yet to come back

- Intel -> ARM was delayed until they could deliver a "big bang" over the whole line up (- MacPro) -> a massive increase in marketshare

- Windows on ARM -> only available on low end, "meh" performance, bad on legacy code -> no hype, no reason for users to switch, no reason for developers to ship ARM builds

Feel free to argue where a 2000 MacOS Intel switch would have landed.

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

@Kronos

Quote:

Kronos wrote: @cdimauro Whatifs are pointless and trying to argue why something that did not happen would have happened isn't much different. But.... Quote: cdimauro wrote: That was different. NextSTEP was already running on different architectures, and even on 68ks. All but the x86 versions had been discontinued years before being bought by Apple.

https://en.wikipedia.org/wiki/NeXTSTEP#History

The last version, 3.3, was released in early 1995, for the Motorola 68000 family based NeXT computers, Intel x86, Sun SPARC, and HP PA-RISC-based systems.
[...]
n 1997, it was updated to 4.4BSD while assimilated into Apple's development of Rhapsody for x86 and Power Macintosh.

So, "years before" = just 2 (TWO) years...
Quote:

- 68k -> PPC "meh" on native code, bad on legacy code -> MS had to bailout Apple a few years later

Except that this move was forced by Motorola decision: Apple was obligated to switch to PowerPCs.
Quote:

- PPC -> Intel (in 2005/6) "meh" for the lowend ( Core Solo), "big bang" for everything above, bad for GPU -> Mac gaming died and has yet to come back

OK
Quote:

- Intel -> ARM was delayed until they could deliver a "big bang" over the whole line up (- MacPro) -> a massive increase in marketshare

OK as well.
Quote:

- Windows on ARM -> only available on low end, "meh" performance, bad on legacy code -> no hype, no reason for users to switch, no reason for developers to ship ARM builds

Wait for the upcoming SoCs from Qualcomm...
Quote:

Feel free to argue where a 2000 MacOS Intel switch would have landed.

I've already did it. And you "just" started writing something else...

@cdimauro

Quote:

cdimauro wrote: I've already did it. And you "just" started writing something else...

Nah, you are just inserting 2005 or even current knowledge into what might have been a 2000 plan/idea.

In 2000 an Intel switched would have been far less of a "big bang" and the future of PPC looked nowhere as bleak as it did in 2004 (when the decision was most likely made).


As for Qualcomm SoCs, kinda like the Moto G5 or the Matrox Parhelia, lots of hype that never transferred into reality. Sure they might be on par with Apples M3, shipping 6 months later but that is nowhere good enough to make the average Windows user/developers consider a switch.

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

I expect less price.
But even 1600 Euro for whole computer is still less
than a1200 with 040 card and pci slots and voodoo3 in 2023 prices.
a1222 will be new and hudred times faster tha 040.

@Kronos

Quote:

Kronos wrote: @cdimauro Quote: cdimauro wrote: I've already did it. And you "just" started writing something else... Nah, you are just inserting 2005 or even current knowledge into what might have been a 2000 plan/idea.

Where? I never mixed the two things: I reported facts belonging to 2000 and other facts belonging 2005, on their respective contexts.

You, instead, have introduced new things belonging on different periods of time and are continuously jumping and mixing stuff, whereas I was always talking and reporting facts of 2000.
Quote:

In 2000 an Intel switched would have been far less of a "big bang" and the future of PPC looked nowhere as bleak as it did in 2004 (when the decision was most likely made).

That's your opinion. However I've reported facts which suggest the opposite.
Quote:

As for Qualcomm SoCs, kinda like the Moto G5 or the Matrox Parhelia, lots of hype that never transferred into reality. Sure they might be on par with Apples M3, shipping 6 months later but that is nowhere good enough to make the average Windows user/developers consider a switch.

Not on par: they distroyed Apple's M3. As well as Intel's and AMD's x86/x64 chips on their market of excellence: desktop.

The used benchmarks aren't cherry-picked ones: they have shown also industry-standard SPEC results.

That's something which very very seriously Apple, Intel and AMD have to look at and propose efficace products to contrast it.

@cdimauro

Quote:

Not on par: they distroyed Apple's M3. As well as Intel's and AMD's x86/x64 chips on their market of excellence: desktop. The used benchmarks aren't cherry-picked ones: they have shown also industry-standard SPEC results.

I'm pretty sure they ran them against M2 a few weeks (or was it just 1) before the M3 were announced.
With a shipping estimate in 2024.

Also EVERY benchmark is cherrypicked especially if it is done by a vendor on yet to be released chips.

On top of that even if they were twice as fast as anything Apple, Intel or AMD have today, still not enough to swing the masses over the Window on ARM.

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

@Kronos

Quote:

Kronos wrote: @cdimauro Quote: Not on par: they distroyed Apple's M3. As well as Intel's and AMD's x86/x64 chips on their market of excellence: desktop. The used benchmarks aren't cherry-picked ones: they have shown also industry-standard SPEC results. I'm pretty sure they ran them against M2 a few weeks (or was it just 1) before the M3 were announced.

Yes, but M3 didn't change a lot for single core performances.
Quote:

With a shipping estimate in 2024.

Which might be close.
Quote:

Also EVERY benchmark is cherrypicked especially if it is done by a vendor on yet to be released chips.

Well, SPEC is... SPEC. And numbers can improve with better compiler support.
Quote:

On top of that even if they were twice as fast as anything Apple, Intel or AMD have today, still not enough to swing the masses over the Window on ARM.

Not yet. It depends more on developers now. But at least the new platform looks promising for them.

@ppcamiga1
Quote:

I expect less price. But even 1600 Euro for whole computer is still less than a1200 with 040 card and pci slots and voodoo3 in 2023 prices. a1222 will be new and hudred times faster tha 040.

That only works with someone thinking of spending big money on A1200 setup.

The other thing is that if Trevor had bit the bullet when the optimum deal was on the table, that €1600 complete A1222 system would have been a €600-700 system.

@cdimauro

Quote:

Well, SPEC is... SPEC.

Which is nice, if your running SPEC for a living

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

@Kronos

Quote:

Kronos wrote: @cdimauro Quote: Well, SPEC is... SPEC. Which is nice, if your running SPEC for a living

SPEC is a very large collection of applications which covers a lot of real-world uses cases / scenarios.

@OlafS25

Quote:

it is certainly faster than 68k (without having the retro feeling) but for what software?

Is that a serious question?

Pretty much any OS4-native audio or video software will benefit from the extra power provided by the PPC. Just off the cuff: Rave, DigiBooster 3, Audio Evolution, MilkyTracker, Protrekkr, Phonolith, ffmpeg, VideoClipper...

Quote:

for some ports of old games and a aging browser who is not updated for years?

If this is what you believe OS4 software boils down to, it's you that have not been updated for years apparently.

_________________
The Rear Window blog

AmigaOne X5000/020 @ 2GHz / 4GB RAM / Radeon RX 560 / ESI Juli@ / AmigaOS 4.1 Final Edition
SAM440ep-flex @ 667MHz / 1GB RAM / Radeon 9250 / AmigaOS 4.1 Final Edition

@Trixie

What are you doing these days with your OS4 box?

@Trixie

Quote:

If this is what you believe OS4 software boils down to, it's you that have not been updated for years apparently.

What a way to pander to your prospective customers! Basically you think we should aleady know about the wonderful world of PPC specific software prior to even hearing about the A1222+!

Newsflash: Most Amigans are more than happy with 68k software and are somewhat perplexed at the PPC/OS4.x offerings TBH!

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

@Rob

Quote:

That only works with someone thinking of spending big money on A1200 setup.

True, but that group of people is probably close to zero.
They are two different markets for two different interests.

Anyone that is planning to purchase one are probably likely to ridicule those wanting to buy the other "you are going to spend that kind of money on that? lol"

So I don't think there are any people that actualy sit there and try to decide "which one should I buy". That would be like asking a Ford-fan what kind of Holden he is going to purchase.


Quote:

The other thing is that if Trevor had bit the bullet when the optimum deal was on the table, that €1600 complete A1222 system would have been a €600-700 system.

If he could have gotten it out in 2016/17 he would probably have been able to sell a thousand units and thus the cost per unit could be low.
Right now I doubt he will even sell a hundred units and the price reflects that.

I wonder what the long holdup could have been. I don't think it is delays for the software, all post-CMB amigas have been basically "buy incomplete system, get updates later (maybe)" so why not have tried to push it as-is back then when the market existed and there was demand.

@Kronos

You're comments imply a narrative in which Steve Jobs, and by extension all of Apple were some PowerPC fan-boys and only switched to intel processors as a last resort.

Companies, especially technology hardware product companies with an R&D budget, maintain multiple streams of research for potential future paths. Meaning that Apple always seriously considered a switch to x86 while also always seriously considering the next generation of PowerPC processors. It's not binary.

Steve Jobs, and by extension Apple, have historically been anti Microsoft, but not anti intel. When Steve Jobs returned to Apple in 1997 he found a legacy of many bad decisions: SKU nightmare, PReP/CHRP clones, the Newton, and yes PowerPC.

NeXT had already moved to x86 because intel/AMD were at the time dominating the price/performance equation. Even Be Inc. capitulated and adopted x86 from 1998. Steve famously went on a clean-up stint as the newly appointed interim CEO of Apple and corrected many of those legacy bad decisions, but the PowerPC ship had already left the port, and many important software partners had already developed PowerPC applications. So in this regard, Steve's hands were tied and he needed to continue with PowerPC, which lets face was near the bottom of his pressing issues.

In typical Steve Jobs fashion, he identified the silver lining in the PowerPC and leaned into them with all the marketing. But in reality, his hardware and software teams were always on the lookout for the best mix that would give Apple the required edge. Internally, they ran many of the server-side apps on x86 while at the same time spending millions on marketing the G4 and G5 Xserve line of products and PowerPC version of Mac OS X Server.
Does it make them hypocrites for not eating their own dog food? Possibly, but at end of the day they're a multi-national business and they owed it to their shareholders to do everything in their power to improve their profitability and market valuation, which kicked off in earnest from 2003.

If the OS X and SDK conditions were right for an x86 switch in 2001, then that's when it would've occurred, but as we know those conditions only ended up prevailing in 2005, and commercial gravity took over from there.

In this regard A-EON, Hyperion, and the MorphOS team are no different. They maintain a PowerPC stance because of the commercial gravity pulling in the direction of existing/completed OS work and developed apps. When it costs them more to maintain PowerPC then to switch away, then the commercially sensible thing to do will be to switch away. Which is kind of where we find ourselves today, it's just that it's happening in Amiga slow motion.

_________________
All the way, with 68k

@ppcamiga1

Quote:

ppcamiga1 wrote: I expect less price. But even 1600 Euro for whole computer is still less than a1200 with 040 card and pci slots and voodoo3 in 2023 prices.

And it's also less than an Apple 1 board sold at a Sotheby's auction, and so what?

The Amiga community has more options than just between an A1222+ and a PCI daughterboard expanded A1200 with 040.
Since we're making up performance numbers, the Apollo V4 Standalone 080 + SAGA is 300 times faster than an 040 and 300% faster at running 68k apps than the A1222+
Since we're picking and choosing hardware for cost comparisons, the Apollo V4SA costs a quarter of an A1222+ with Radeon RX580 in a Corsair Obsidian 1000D case with custom water-cooling loop.

Since the A1222+ is not a real Amiga, many Amigans prefer the performance advantages of a real Amiga 1200 with PiStorm32 accelerator. 1000 times faster than 040.

And if one is buying non-Amigas, then a Power Mac G5 running MorphOS is half the price of A1222+ and 5000 times faster than 040. And it's PowerPC big-endian.

Last edited by agami on 13-Nov-2023 at 06:53 AM.

_________________
All the way, with 68k

@all
With all the talk about the disappointing PPC G5 that sank the PPC desktop market and the disappointing PPC Cell processor which sank the console market, I just did some 7-zip benchmark comparisons of these cores.

https://amigaworld.net/modules/newbb/viewtopic.php?topic_id=45079&forum=17#864827

IBM soiled their reputation with these cores and killed PPC as much as Apple leaving the AIM alliance for x86.

@cdimauro

Quote:

OK, but here you were talking about IBM's G5, right? Not the Motorola's one.

Yes, the G5s Apple put in the final 64 bit PPC Macs.

@ppcamiga1

Quote:

I expect less price.

We all did. Timing is the worse.

Quote:

But even 1600 Euro for whole computer is still less than a1200 with 040 card and pci slots and voodoo3 in 2023 prices. a1222 will be new and hudred times faster tha 040.

That may be so. Just like my A1/XE setup cost me $2,000 against $2,500 for my overall A1200 setup and made my A1200 look like crap. But, people will say it's not a real Amiga, compare it with their PC or Mac, then say it's too expensive. Amiga people have been saying the same thing the last 20 years. Sometimes I wonder if Amiga users were ever Amiga users because they always say how good the PC and Windows is now and rely on Windows to install Amiga software. On their Amiga. Without complaining that their real Amiga and Amiga cards cost so much more than a PC and is so much slower.