[A-EON updates] The "Golden" Child - Nemo revision 2.1

[commodorejohn]

The X1000 is standard PC components on a PCB inside a *generic PC case* sold at a *ridiculous high price* to brand name followers.

I'm no X1000 supporter, but I missed the part where a PowerPC CPU is "standard PC components..."

I'll give $10 to the first person to run Dhrystone on one of these; I want to see how my $375 4x2.5GHz G5 stacks up

[commodorejohn]

Every day I say a prayer that the port to x86 or at least ARM. Staying on dead end power pc is such a waste of a perfectly nice (or will be eventually) OS.

I don't have anything against the idea of an Intel or ARM port (as really, I'm not that invested in NG Amigoid OSes,) but I think it's a fallacy to think that PPC is a "dead end" while x86 is going to reign supreme forever. Architectures come and go, wax and wane, often multiple times before they die (if they ever really do.)

Hell, PPC itself was a rising star back in the early '90s, when Intel CPUs were clunky and terrible and RISC was going to be the Next Big Thing that would sweep away everything else forever. (IBM even tried to make that happen with a line of 603-based workstations and laptops, but it seems they forgot that they hadn't actually been the PC-industry leader since they invented a new bus and demanded that people pay them for the privilege of switching to it.) It's still not dead now, but that prophesied revolution certainly never happened.

x86/x64 has had a long and successful run and doesn't appear to be going anywhere for the moment, but this too shall pass. The real goal ought to be to have a portable OS, not to find the most popular architecture of the moment and hitch your wagon to that thinking it's always going to be there for you.

(And anyway, I don't believe PPC is a dead end, not while there's still plenty of inexpensive PPC hardware out there for the taking, much of which is quite good. Apple's made some shoddy machines for the low-end markets, but their high-end machines have been built to last, all the way back to the Mac II. I've just saved three G4 Power Macs from the recycle center, and for all the battering they've taken and as old as they are, they still run like new!)

[commodorejohn]

Well, it was exactly that during longer than a decade, and the fact is that this very CPU was designed for potential use in year 2007 level Apple laptops, so you are wrong.

Well, if you're going to define "standard PC" to include Apple's pre-Intel Macs, I suppose, but I don't know of anyone else who extends the definition that far... :/

But in a way you are right; the PPC doesn't even qualify to neither "PC" nor "Desktop" anymore, that era ended half a decade ago (today PPC is used for routers, printers, cars, etc). And if this is your criteria for "being Amiga", then there is something fundamentally wrong.

I actually said nothing at all about what makes an Amiga. And that's not how I feel about it anyway; when it comes to the Amiga I'm a 68k man all the way. I just don't think that PPC is a "dead end," or that i7 is the Second Coming and of Intel's kingdom there shall be no end.

But just because PPC has fallen out of favor currently (though if you haven't noticed, RISC architectures have been making quite a bit of a comeback, when you factor in smartphones and other mobile devices, and PPC specifically is doing all right, what with driving all three of the current-gen game consoles,) that doesn't mean that it doesn't have a future. The industry moves in unpredictable ways, and it may be that someone's going to come along and do for the PPC architecture what Core 2 did for Intel, pulling it out of its slump and making it a major player again. Or maybe not. Who can say? Fate is fickle.

I do agree that the X1000 is overpriced and underpowered (as I've said multiple times, I'm not an A-eon defender, and I certainly have better uses for $2500 than buying something that probably isn't as good as my G5 Power Mac.) I just don't think it's fair to tie the X1000 and the PowerPC architecture as a whole up in one inextricable bundle.

[commodorejohn]

The biggest problem with PPC macs is probably the software. You have to admit when Apple was shoveling out G3 and G4 macs and advertising that they were the fastest thing ever seen, they ran like dogs compared to midrange PCs running win XP. At the end of the day, that's all that matters. For whatever reason their software stack did not do it any justice. Compare two identical PPC Macs, one running linux and one running OS X. OS X trundles along while linux flies.

Now compare two Intel macs, one with OS X and one with Linux. The performance is similar. If PPC was so great, why was Apple never able to truly take advantage of it?

Yes and no. Apple's rush to OS X was definitely a problem, as even their early G5 machines weren't all that well-suited to it - and they were putting it on mid-range G4s running at half or less the speed of even the low-end Power Mac G5! But if you put OS9 on the same machine (assuming it's early enough in the G4 cycle that they hadn't dropped support entirely,) there's a marked performance increase.

The Power Mac G4 I rescued from the recycle center (single-CPU 533MHz,) for example, is the most responsive thing I've run classic Mac OS on yet, and that includes modern computers running SheepShaver. With the exception of large webpage rendering (which would probably see a marked improvement if I had something better than the piddly 16MB video card they shipped it with,) it comes in only somewhat under my old 1.2GHz P4 laptop running XP, and absolutely thrashes my old 800MHz Celeron WinMe box.

I think the problem is less that Apple's software is bad and more that they always insist on everyone using their latest and greatest, hardware suitability be damned. They even staged a "funeral" for OS9 when it's a perfectly workable OS (for its time) and far better suited to the hardware they were shipping when they introduced OSX.

[commodorejohn]

The CPU architecture have nothing to do with the hardware quality.

True. But it is pretty grody, being a 64-bit expansion to a 32-bit update to a 16-bit remake of an 8-bit CPU that frankly wasn't all that great to begin with. I'm not going to claim that all hardware should be chosen based on design prettiness, but it's like what they say about knowing what goes into sausage; having tinkered around under the hood of x86 systems, I'll never look at one the same way again.

[commodorejohn]

So how much does it cost just for the CPUs themselves and nothing else? I'm guessing maybe $1000 for the pair? Just remember that the Sams come with only a single PPC whereas the X1000 comes as a pair thus explaining such a high cost...

It's a dual-core CPU, not two CPUs - quite a bit cheaper of an approach. I'd suspect more of the price is in paying off the development cost and the higher per-unit cost of a small production run versus a large one, though you'd have to ask A-eon to know for sure.

[commodorejohn]

True, but x86 is boring as hell.

There's gotta be a cheaper way to get PPC hardware, really there does. Nintendo can sell a Wii for $150. Granted that's only a souped-up G3 in there, but it still can't be that expensive to manufacture the PA6T...

[commodorejohn]

And in the case of the X1000, you've also got to factor in low production volume into the price.

Well naturally - I was just a bit stunned by the cost of the CPU...

[commodorejohn]

No, BeOS died because it had no software. Simple.

Probably both, actually - no software because all the support was for Windows, couldn't get support because it couldn't compete with Windows because it had no software. It's a vicious cycle that affects a lot of nascent platforms if they don't have a real killer app. Hell, even the IBM PC languished for well over a year until Lotus 1-2-3 gave people an incentive to look into it.

Actually, BeOS did and does have a  comparable selection of software. Emulators, video editors, video  players, e-mail, modern browsers. They even have Firefox, Thunderbird,  and other software we continue to lack.

Compared to OS4, yes, but OS4 was a niche market from the get-go. The relevant comparison for BeOS at the time was Windows or Mac OS, both of which had far greater commercial support (and open-source development didn't really take off until the end of BeOS's lifespan.)