The Borg have sent a gift to Earth
July 18, 2000 3:18 PM Subscribe
The Borg have sent a gift to Earth The much rumored Apple cube seems to be real. Apple Insiders has just posted what looks to be a promotional photo of the new computer. Looks pretty cool. Not sure why there are vents on top but still, I wouldn't mind having one.
You probably could fit a Pop Tart into the CD-ROM drive.
posted by tomalak at 5:17 PM on July 18, 2000
posted by tomalak at 5:17 PM on July 18, 2000
Ack! They took the pic off due to "threat of Apple legal action" do you happen to have the pic cached, and can you mirror it somewhere?
posted by deckard at 8:40 PM on July 18, 2000
posted by deckard at 8:40 PM on July 18, 2000
It'll be shown legally at the show tomorrow, probably. No worries.
posted by endquote at 9:20 PM on July 18, 2000
posted by endquote at 9:20 PM on July 18, 2000
I assumed it would be a big server thing. So much for rumors.
posted by iamcal at 11:37 AM on July 19, 2000
posted by iamcal at 11:37 AM on July 19, 2000
Am I the only one who thinks that the cube looks like an executive trashcan? I'd almost expect a foot pedal on the front to pop the top open. As a long-time PC user who admired the iMac, G3/4, and iBook, I think this gizmo may finally cure my Mac envy.
posted by harmful at 12:10 PM on July 19, 2000
posted by harmful at 12:10 PM on July 19, 2000
It's so sad that all Apple can offer anymore is different colors and boxes for the same thing.
posted by john at 12:33 PM on July 19, 2000
posted by john at 12:33 PM on July 19, 2000
While I don't know that the PowerMac Cube is anything special other than innovative packaging (which, admittedly, did a lot for the iMac), I thought it was impressive that they are updating the current PowerMac models with standard gigabit ethernet and dual processors on the 450 and 500 MHz models for the same price.
And as for doing something different, OS X is going to be different than most anything before. Will it be a success? I don't know, but it is different.
posted by daveadams at 1:17 PM on July 19, 2000
And as for doing something different, OS X is going to be different than most anything before. Will it be a success? I don't know, but it is different.
posted by daveadams at 1:17 PM on July 19, 2000
Have you seen the prices??? Ye Gods, when will Apple produce a computer which is actually cost competitive?
Courtesy of JC's PC News'n'Links:
400MHz - 1 CPU [$1599]
450MHz - 2 CPU ***standard*** [$2499]
500MHz - 2 CPU ***standard*** [$3499]
They can't be serious!! (And topping out at 500 MHz is crippled, simply crippled. But then, that's the best Moto can deliver, because it's been becoming obvious that Moto no longer really cares all that much.)
For $1600 you can put together a computer now with a 1 GHZ Athlon, and I don't want to hear any myths about "PPC is 3 times faster per megahz" because that's baloney. And what kind of PC can you build these days for $3500?
In its entire history, Apple has never delivered a computer which was price competitive with contemporary PCs.
I guess you're supposed to be willing to pay a considerable surcharge for the indescribable pleasure of running MacOS.
Hmmmph.
posted by Steven Den Beste at 1:31 PM on July 19, 2000
Courtesy of JC's PC News'n'Links:
400MHz - 1 CPU [$1599]
450MHz - 2 CPU ***standard*** [$2499]
500MHz - 2 CPU ***standard*** [$3499]
They can't be serious!! (And topping out at 500 MHz is crippled, simply crippled. But then, that's the best Moto can deliver, because it's been becoming obvious that Moto no longer really cares all that much.)
For $1600 you can put together a computer now with a 1 GHZ Athlon, and I don't want to hear any myths about "PPC is 3 times faster per megahz" because that's baloney. And what kind of PC can you build these days for $3500?
In its entire history, Apple has never delivered a computer which was price competitive with contemporary PCs.
I guess you're supposed to be willing to pay a considerable surcharge for the indescribable pleasure of running MacOS.
Hmmmph.
posted by Steven Den Beste at 1:31 PM on July 19, 2000
I don't want to hear any myths about "PPC is 3 times faster per megahz" because that's baloney.
Are you sure? I mean, I don't really believe all of Apple's claims at face value, but surely you realize it's possible for a processor to be faster, even at lower clock speeds. There are lots of other factors involved in determining processor speed, most notably including instructions per cycle. Surely the 3.5 gigaflop number (seven times the clock speed!) should mean something. What about system bus speeds, cache speed/size?
Who knows, maybe Apple really is still producing processors that are slower than anything you can even still purchase in the PC world. But my dual-500MHz Pentium III at work is still a lot faster than anything I need... for now.
posted by daveadams at 2:05 PM on July 19, 2000
Are you sure? I mean, I don't really believe all of Apple's claims at face value, but surely you realize it's possible for a processor to be faster, even at lower clock speeds. There are lots of other factors involved in determining processor speed, most notably including instructions per cycle. Surely the 3.5 gigaflop number (seven times the clock speed!) should mean something. What about system bus speeds, cache speed/size?
Who knows, maybe Apple really is still producing processors that are slower than anything you can even still purchase in the PC world. But my dual-500MHz Pentium III at work is still a lot faster than anything I need... for now.
posted by daveadams at 2:05 PM on July 19, 2000
Yes I'm sure that that particular statistic is baloney.
The reason I'm sure is that it's based on a specifically cooked benchmark which was designed to take advantage of a particular Altivec instruction. Howver, the instruction in question is not one which comes up much in normal application coding -- indeed, Altivec itself doesn't come up much in normal application coding.
But in the real world, when you take applications which have been coded for both (PhotoShop, Word, Excel) you tend to find that MHz for MHz they perform about the same as PCs do. (On specific operations, sometimes the Mac will be faster and sometimes the PC will be. Certain Photoshop operations are particularly fast on the Mac because they've been rewritten to take advantage of said Altivec instructions. However, those are very special cases.) And that's the only place where performance really matters: in the real world, running real apps.
None of the cooked benchmarks which show the PPC to be vastly more powerful is even remotely representative of normal code. Every one of them is specifically cooked to show the PPC in its best light and the x86 in its worst.
In the real world, running real apps, the PPC tends to be approximately comparable to an x86 of the same speed.
The simple fact is that that 400 MHz PPC is completely outclassed by the current generation of PCs and it's too damned expensive for what it's providing.
posted by Steven Den Beste at 3:22 PM on July 19, 2000
The reason I'm sure is that it's based on a specifically cooked benchmark which was designed to take advantage of a particular Altivec instruction. Howver, the instruction in question is not one which comes up much in normal application coding -- indeed, Altivec itself doesn't come up much in normal application coding.
But in the real world, when you take applications which have been coded for both (PhotoShop, Word, Excel) you tend to find that MHz for MHz they perform about the same as PCs do. (On specific operations, sometimes the Mac will be faster and sometimes the PC will be. Certain Photoshop operations are particularly fast on the Mac because they've been rewritten to take advantage of said Altivec instructions. However, those are very special cases.) And that's the only place where performance really matters: in the real world, running real apps.
None of the cooked benchmarks which show the PPC to be vastly more powerful is even remotely representative of normal code. Every one of them is specifically cooked to show the PPC in its best light and the x86 in its worst.
In the real world, running real apps, the PPC tends to be approximately comparable to an x86 of the same speed.
The simple fact is that that 400 MHz PPC is completely outclassed by the current generation of PCs and it's too damned expensive for what it's providing.
posted by Steven Den Beste at 3:22 PM on July 19, 2000
Last (wholly unsubstantiated) rumour I heard had IBM jumping in to produce 600-700 MHz PowerPC chips. Given Motorola's breathtaking failure to execute over the last year, it might even be true.
posted by snarkout at 3:32 PM on July 19, 2000
posted by snarkout at 3:32 PM on July 19, 2000
Read the section labelled "MacWorld 2K: Like fish in a barrel".
Among other blunders, Jobs claims to have issued the first SMP personal computer -- which comes as a big surprise to certain PC users who have been using them for months.
posted by Steven Den Beste at 3:39 PM on July 19, 2000
Among other blunders, Jobs claims to have issued the first SMP personal computer -- which comes as a big surprise to certain PC users who have been using them for months.
posted by Steven Den Beste at 3:39 PM on July 19, 2000
For me, style will never be more important than the price vs. performance ratio. I'm disappointed with Apple's 'we can't sell a cost effective computer but at least our boxes look cool' attitude. Function is more important than style.
posted by Mr. skullhead at 6:23 PM on July 19, 2000
posted by Mr. skullhead at 6:23 PM on July 19, 2000
There was a version of the Power Mac 9500 a couple years ago that came with dual processors. It didn't do much, since software had to be specifically written for it or else the second chip just sat there idle, but it existed.
As for "Tech-Report," it's hard to take him seriously when he doesn't even realize that the "Velocity Engine" isn't some toy word Apple made up to make the Power Mac sound cooler; it's a specific part of the G4 chip's design that allows it to process data in 128-bit chunks, which is 2 to 4 times faster than other chips. And yeah, that actually speeds stuff up, believe it or not.
And while I'm not sure what benchmarks Apple used in the keynote today to plug their new machines, the best G4s do beat the best PCs in Intel's own testing benchmarks. Even Adobe says the G4s run Photoshop much faster than any other chip. And that is the sort of thing that Apple's core market is looking for.
Of course, after a certain point the whole speed argument is a bit specious anyway; PCs and Macs are different machines that don't run the same software. If you're just gonna do basic word processing and web surfing, sure, blowing the $ for a high-end G4 is unnecessary. (But then, the new base-level iMac is only $800, heh.) If you're a hardcore graphics designer, these things are preferable in most cases.
posted by aaron at 6:57 PM on July 19, 2000
If you're a hard-core graphics designer, in this day and age you're going to want nothing whatever to do with the ATI RAGE 128. That chip is three generations behind what current PCs are using; in fact, it's so primitive that it isn't even offered for sale in the PC market. (Try comparing it to the NVidia GeForce 2 GTS I just purchased; it's a joke by comparison.)
I also spotted another interesting spec: that G400 single-processor system comes with 64M of PCI100 memory; haven't they heard of PCI133 down there California way?
In point of fact, technologically these suckers are completely behind the times: old wine in new bottles. It's just the same old crap repackaged in a pretty box.
(By the way, just how the heck are you going to use SMP with MacOS9 when threads are so crippled that no application programmer uses them? One of your CPU's is going to spend a lot of time waiting, while the other one does all the work due to the app developer creating his own round-robin multitasker inside his app to take the place of threads. As a PC user, one of my big bitches with Poser is that it's implemented this way because that's the only way they could make it work on the Mac, and when they ported it to Windows they didn't bother rewriting to take advantage of threading which DOES work correctly under Windows. So it continues to do its own multitasking, and it does it really badly. If you run Poser on an SMP Mac, it's going to run exclusively on a single CPU. And I think you'll find that a lot of other apps are like that. Because of the MacOS legacy, SMP right now is virtually useless for single applicatios; at best it permits you to run two apps simultaneously. Oh, and if Apple just invened SMP, howcome it's in Win2K which was released a couple of months ago?)
Feh.
posted by Steven Den Beste at 7:17 PM on July 19, 2000
I also spotted another interesting spec: that G400 single-processor system comes with 64M of PCI100 memory; haven't they heard of PCI133 down there California way?
In point of fact, technologically these suckers are completely behind the times: old wine in new bottles. It's just the same old crap repackaged in a pretty box.
(By the way, just how the heck are you going to use SMP with MacOS9 when threads are so crippled that no application programmer uses them? One of your CPU's is going to spend a lot of time waiting, while the other one does all the work due to the app developer creating his own round-robin multitasker inside his app to take the place of threads. As a PC user, one of my big bitches with Poser is that it's implemented this way because that's the only way they could make it work on the Mac, and when they ported it to Windows they didn't bother rewriting to take advantage of threading which DOES work correctly under Windows. So it continues to do its own multitasking, and it does it really badly. If you run Poser on an SMP Mac, it's going to run exclusively on a single CPU. And I think you'll find that a lot of other apps are like that. Because of the MacOS legacy, SMP right now is virtually useless for single applicatios; at best it permits you to run two apps simultaneously. Oh, and if Apple just invened SMP, howcome it's in Win2K which was released a couple of months ago?)
Feh.
posted by Steven Den Beste at 7:17 PM on July 19, 2000
Jobs, for better or worse, chooses his words carefully. He didn't say it was the first SMP machine, he said it was the first one to come "standard" (i.e., as the default on a company's high-end machine, I'm guessing). Of course in the Wintel world once you get into serious, high-end machines there's not really such a thing as a "model" in the same way Macs or cars come - they're often put together using hand-picked components.
But anyhow - for me it doesn't matter, I still want a cube. And I'll probably get one if I juggle my finances just right.
posted by mikel at 7:30 PM on July 19, 2000
But anyhow - for me it doesn't matter, I still want a cube. And I'll probably get one if I juggle my finances just right.
posted by mikel at 7:30 PM on July 19, 2000
>>just how the heck are you going to use SMP with MacOS9...?
You're not; it's for OS X.
I'll take your word for it WRT the ATI chip, but if the Photoshop-type programs still perform better even with it, what's the diff?
posted by aaron at 7:50 PM on July 19, 2000
But you're only going to use SMP in a big way under OSX once the majority of the Mac software being offered for sale has been specifically written for OSX, which by definition means it won't work on anything earlier.
In other words, a transition equivalent to the Win16 to Win32 transition in the Windows world at the applications level. That transition took four years to complete, to reach the point where substantially no Windows software was being offered for sale which would run on Win 3.1.
The Mac will have to reach the same point, where substantially none of the software being offered for sale will work on MacOS9 or earlier, and I see that as being a very slow transition (possibly five years -- the economics are different). In the mean time, the utility of SMP in a Mac will be extremely limited, and those dual-CPU boxes will effectively run most of the time at speeds comparable to a single-CPU box.
(Damn, I promised myself I wouldn't get involved further in this thread. I do want to make this comment: I will have no respect for the Mac until that transition I described above has actually happened, where essentially everything is Mac OSX or later including the apps and where no-one uses MacOS9 any longer because they can't buy any software for it (in other words, when MacOS9 joins Win 3.1 in the bit bucket because its API is no longer used). Then and only then will the Mac be anything remotely like a modern computer -- assuming they actually do kiss off ATI and make that deal with NVidia to give them a reasonable graphics card for a change. OK, this really is my last post in this thread. I promise. Cross my heart.)
posted by Steven Den Beste at 8:31 PM on July 19, 2000
In other words, a transition equivalent to the Win16 to Win32 transition in the Windows world at the applications level. That transition took four years to complete, to reach the point where substantially no Windows software was being offered for sale which would run on Win 3.1.
The Mac will have to reach the same point, where substantially none of the software being offered for sale will work on MacOS9 or earlier, and I see that as being a very slow transition (possibly five years -- the economics are different). In the mean time, the utility of SMP in a Mac will be extremely limited, and those dual-CPU boxes will effectively run most of the time at speeds comparable to a single-CPU box.
(Damn, I promised myself I wouldn't get involved further in this thread. I do want to make this comment: I will have no respect for the Mac until that transition I described above has actually happened, where essentially everything is Mac OSX or later including the apps and where no-one uses MacOS9 any longer because they can't buy any software for it (in other words, when MacOS9 joins Win 3.1 in the bit bucket because its API is no longer used). Then and only then will the Mac be anything remotely like a modern computer -- assuming they actually do kiss off ATI and make that deal with NVidia to give them a reasonable graphics card for a change. OK, this really is my last post in this thread. I promise. Cross my heart.)
posted by Steven Den Beste at 8:31 PM on July 19, 2000
Oh, OK. Steven doesn't like Macs. Check.
I'm not a Mac guy either, in the sense that I haven't ever owned one, though I certainly know my way around one (I've networked more than a few). It should be bleeding obvious that the reason a PPC Mac outperforms Wintel even at "half" the clock speed is that Wintel is still supporting ridiculously outdated x86 processor code.
But enough about that. Can we discuss who's going to buy this thing? I might like one, although it does appear a bit pricey compared to the regular G4. Is it really just for graphic designers and executives?
posted by dhartung at 11:38 AM on July 20, 2000
I'm not a Mac guy either, in the sense that I haven't ever owned one, though I certainly know my way around one (I've networked more than a few). It should be bleeding obvious that the reason a PPC Mac outperforms Wintel even at "half" the clock speed is that Wintel is still supporting ridiculously outdated x86 processor code.
But enough about that. Can we discuss who's going to buy this thing? I might like one, although it does appear a bit pricey compared to the regular G4. Is it really just for graphic designers and executives?
posted by dhartung at 11:38 AM on July 20, 2000
« Older Breakin' dah law breakin' dah law! | Is this the strangest contest ever? Newer »
This thread has been archived and is closed to new comments
posted by greyscale at 4:20 PM on July 18, 2000