Slashdot Mirror
Apple Marketing Hypes New PowerMacs
Wacky_Wookie was only one of many who wrote in with a mention of Apple's "leak" of specifications for a new line of PowerMacs to be dubbed "G5", apparently running the new PowerPC 970 CPUs. No offense, but anyone who thinks it was a mistake or leak doesn't understand marketing. :) Update by J : In case those linked sites get taken down too, try
MacNN.
But you can get it from here.
Also more on the story here.
My journal has hot
The was from a graphic directly from store.apple.com on the PowerMac site; not from a little teen's weblog (heh)...the credibility is high, and all of the sites that had posted the information have since removed the graphic per request from apple. macrumors, macnn, appleinsider, mac minute, etc.
the posted specs went far beyond the expectations of...anyone.
* 1.6Ghz, 1.8Ghz, or dual 2Ghz PowerPC G5 Processors
* Up to 1 Ghz processor bus (!!)
* Up to 8 GB of DDR SDRAM
* Fast Serial ATA hard drives
* AGP 8X Pro graphics options from NVIDIA or ATI
* Three PCI or PCI-X expansion slots
* Three USB 2.0 ports
* One FireWire 800, two FireWire 400 ports
* Bluetooth & Airport Extreme ready
* Optical and analog audio in and out
Quite a leap from the current dual 1.42Ghz G4 boxes, with a 166Mhz bus...
four nine eighteen twenty-7 thirty-nine forty-7 fiftyeight sixty-nine seventy-9 eighty-8 one-hundred-and-nine one-twenty
here, here and here.
Going back through my memory, I can think of a couple other very important leaks:
* ATI leaks news of updated PowerMac G4s (and summarily gets poor product placement in them afterwards, with Apple favoring Nvidia for years to come)
* Time Canada posts story of new flat-panel iMac before paper issue even hits the streets.
I'm sure there have been more, but those seem to be the most important.
This one will go down as probably the most significant leak in quite a while.
Otherwise, with those specs, it's about time. When it does come out, it will at least give Intel/AMD a run for the money (remember, the 64 bit stuff will be out by then), instead of eating their dust.
Well, how many PPC processors can support 8 GB of RAM with a FSB of half the chip's frequency ? :)
a lot of G4's shipped with just. plain. awful sound cards. this is welcome news.
It's not a G5, it's a PPC970, completely different beasts.
Newsflash, kiddo: neither Motorola nor IBM sell a CPU called the "G4". "G4" was a "marchitechture" term coined by Apple in the spirit of Motorola's internal "G3" codename for the PPC750. The chip inside any "PowerMac G4" is some flavor of a Motorola PowerPC 7400, no matter what Apple calls it.
You can pretty much bet the farm that Apple will call every varient of the PPC970 they ship a "PowerPC G5".
1GHz bus? gimme a break. Intel hasn't yet reached this. Two points impossible.
Ahem. ("1ghz" is probably apple marketing-speak, but it's always been known that the PPC970 will have a stupidly fast FSB -- Intel isn't the only company that can innovate in this field, eh?)
Almost believable, but for the moment Apple are phasing out the use of NVIDIA cards in their machines.
Simply and 100% wrong. Apple has been doing pretty much exactly the same thing for the last three years on this front: providing whichever of the two offered them the best OEM pricing as the default configuration, and offering the other as a build-to-order option. They will continue to do this.
Also, Apple have a long standing habit of using Firewire instead of USB 2.0
Here, you may be correct, but there are two issues that may force them to start shipping "USB 2.0" connectors: first, the USB consortium has recently declared that all USB ports are "USB 2.0" (yes, this is weird and stupid), and secondly it's actually getting a bit difficult to source USB controllers that only support the 1.0/1.1 specs.
Once again use of the verbal "One" instead of the numeric. Only one FW800 port? Why would Apple stick with FireWire 400 anyway? I mark this impossible
FW400 and FW800 use different connectors, and there are not yet many FW800 products on the market. This is called "covering your bets" and "not pissing off your customers". BTW, 1x FW800 and 2x FW400 is also the configuration on the 17" AlBook, so they've already shipped one machine in exactly this "impossible" configuration.
optical audio in a graphics machine? I'm sorry but this sounds like wishful thinking.
No, it sounds like you have no idea what you're talking about. Do you have any idea how many macs are used in audio production? Are you aware that Apple sells their own high-end audio composition program? The only surprise about a PowerMac with optical TOSlink is that they didn't do it years ago.
News for Nerds. Stuff that Matters? Like hell.
Although I agree that some of the wording is not consistant with Apple's regular product descriptions, the specs are not at all impossible. If you've been keeping up with the news behind the G5, you'll know a few things. First, the G5 is the PPC970. The "G" product name has a shit-load of product recognition. Like the Pentiums. How close is a Pentium4 to a Pentium3? Not very much. Also, about the system bus, they are using Hyper-Transport which does allow for a 1GHz system bus. I've seen system design specs that indicate a doulbe (or possibly even quad) pumped, bidirectional bus (500MHz out / 500MHz in). About the FireWire 800/400 issues: it's all about cost and bandwidth. Anything more than 1 FireWire 800 port is so unnecessary, it's rediculous. Name anything that can use all that bandwidth, let alone the system bus! All that bandwidth room requires computing power which requires more circuits which, in turn, requires more money. FireWire 400 is fast enough for anything right now. Also, about USB 2.0, look at the number of products out now which requires USB 2.0. Apple would be stupid not to build in support for it. One more thing: AGP 8x is necessary now for graphics.
Totally Life!
ALL replies
A relevant picture can still be found on an Akamai Mirror (taken from an write-up here
Do you understand what he's saying? He's saying that they already had some images done up and perhaps he just copied the wrong one.. I'm sure they have the site ready in staging for Monday
It's not a G5, it's a PPC970, completely different beasts. Not to mention neither Motorola or IBM have 2GHz chips in their roadmap until 2005. Bzzzt One point impossible
Bzzt. Wrong.
G3 and G4 are both Apple marketing terms, not CPU specifications from IBM or Moto.
IBM was expecting low yields of the 2ghz chip but it was always on the roadmap for this year... or have you not been paying attention?
I'm writing this on a Powerbook G4, not a Powerbook 7450 (PPC 7450 is Motos term for the cpu running this beastie).
1GHz bus? gimme a break. Intel hasn't yet reached this. Two points impossible
Bzzt. Wrong. One word for you:
Hypertransport.org
The FSB runs at half the clockspeed of the CPU. A dual 2ghz 970 would have FSB of 1ghz.
The rest of Apples site would say "3 USB Ports" not "Three". Also, Apple have a long standing habit of using Firewire instead of USB 2.0. I take this as one point impossible
The current G4s ship with USB 2.0 chipsets. Firewire and USB2 are NOT in competition. THey have different applications. If you don't belive me, then I ask you to point to a USB2 uncompressed SDI interface? Oh. YOu can't? Shit.
Once again use of the verbal "One" instead of the numeric. Only one FW800 port? Why would Apple stick with FireWire 400 anyway? I mark this impossible
My dual 1.4ghz G4 has a FW800 port and two FW400s. FW800 is a different physical interface than FW400. I'm sure the chipset is also slightlly more expensive.
Bad grammar, but optical audio in a graphics machine? I'm sorry but this sounds like wishful thinking. One more point impossible.
Marketing has always made a habit of playing with grammar.
Macs are not just graphics machine. In fact, the dual 1.4ghz mentioned above is primarily an audio workstation. It has digital audio out already on board.
Optical in/out is a surprise, but not unlikely - it has its advantages.
This is accurate. Like it or not. Apple is back in town where it belongs - on the top.
Wow let's take this one point at a time...
It's not a G5, it's a PPC970, completely different beasts. Not to mention neither Motorola or IBM have 2GHz chips in their roadmap until 2005. Bzzzt One point impossible
Errr... From a marketing perspective G5 is better then 970, it's also consistant with how Apple have named PPC processors in the past, so Bzzzt minnus one for you Plus one for Apple.
1GHz bus? gimme a break. Intel hasn't yet reached this. Two points impossible
Wow! I don't know where to begin with this one. If you really understood anything you were talking about you would know two things: 1. The 970 bus speed runs 1/2 the speed of the processor, so for a 2Ghz Processor, 1 Ghz sounds about right. 2. WTF does Intel have to do with anything? Apple has worked with AMD on the Hypertrasport BUS which should (and apparently does) toast Intel (And Intel has been slower then most with BUS speed lately anyway, even little VIA who have a fraction of the budget of Intel/AMD/ or Apple
The rest of Apples site would say "3 USB Ports" not "Three". Also, Apple have a long standing habit of using Firewire instead of USB 2.0. I take this as one point impossible
Apple was first with USB, and many many many Apple peripherals use USB (i.e. Keyboard and Mouse to name a few important ones). The cost of USB 1.1 vs 2.0 is about nothing so it's a no brainer to use USB 2.0. Also Since you seem to have grammatical issues the image (see below) "three" is grammatically correct "3" isn't.
Almost believable, but for the moment Apple are phasing out the use of NVIDIA cards in their machines. I highly doubt they'll be used. Half a point impossible
Apple isn't phasing out Nvidia, Have you read the specs for the newest Powerbooks? They use Nvidia instead of tradtional ATI. Apple has offered choice in desktop G4 systems for awhile. (It's Microsoft who are phasing out Nvidia in the next XBox)
Once again use of the verbal "One" instead of the numeric. Only one FW800 port? Why would Apple stick with FireWire 400 anyway? I mark this impossible
You and your verbal numbers... get a life. I don't know where you are coming from on this one. The Apple spes are perfectly logical and similar to the latest 17" PowerBooks
Bad grammar, but optical audio in a graphics machine? I'm sorry but this sounds like wishful thinking. One more point impossible
In a graphics machine? Well they are great at graphics, so I'll give you that, but have you ever heard of Digidesign ProTools? How about Apple's own Logic? How about the fact Apple has an entire segment devoted to Music right off there home page (and only part of it it the music store and iPod)? Apple is doing Music big time, and at a very high professional level, optical audio is an unexpected, but very logical addition to the G5
Hmm... there go your 4 1/2 half impossible points (or whatever? heck this is almost desperate... are you the person from Apple who accidently posted this in the first place and this is a lame attempt to counteract the damage before Steve fires your ass?)
--- Nothing To See Here ---
Bad grammar, but optical audio in a graphics machine? I'm sorry but this sounds like wishful thinking. One more point impossible.
This one does puzzle me slightly, but Apple don't consider their machines "graphics machines". Macs are all-purpose. It's no fault of Apple if they are pigeonholed by others as "only for graphics". Also, even if they were "graphics machines," video editing kind of requires sound..."
Not only that, but unless you've been living in a hole, you'd know that within the last year, Apple has purchased Logic. This computer would compliment that nicely.
today is spelling optional day.
Well, ok, maybe in the bigger consumer arean but I've done some pro studio audio and up until the past few years, if you wanted high end quality audio, you did it on a Mac. ProTools et.al, all debuted on the Mac. In fact, my friends studio - of which over the years we tried converting to Windoze multitrack - went back to a Mac and churned out better sounding projects.
There are many uses in non-consumer audio for optical i/o. I mean, shit, my cheap little Shuttle Spacewalker has a coinnector card for optical audio. For them to include it now is obviously smart since we know they're also targeting more musicians (specifically dj's) now as well as graphics people.
Frankly, this guys 'impossible' post is a lot of bullshit for many reasons listed above and throughout.
I now must find a way to ditch my PB G4 1ghz. G5, here I come! *drools*
FireWire 400. Why stick with this? Good question. I think I agree with you here since the 800 ports would be backwards compatible with any 400 device. 1 point impossible
Actually, I can see why they'd want to keep 400s. The 800 connector is different from the standard 400 cable, and even with the availability of adapters to convert from 400->800, it's still easier for them to simply include 400s on the computer so people who buy it can plug their devices in as soon as they assemble the computer, rather than having to go out and buy a special cable just to use their cameras, etc.
Want Slashdot headlines on your site? Try SlashHead
Apple machines have held a performance lead over MS PCs at various times in the past. The new dual 2Ghz machine seems to have a reasonable chance of being one of the fastest desktop machines available - although Apple probably won't be able to hold that lead when the Athlon 64 vs Intel battle hots up again.
I can't wait to grab one of these new Apples - good work Apple & IBM, I say!
That was classic intercourse!
All Macs have been CHRP-compliant since CHRP(and before that, PPRP*) became a standard.
The only difference is that the MacOS(OSX included) is not CHRP-compliant. It requires a Mac ROM(or modern equivalent). LinuxPPC(RIP) has been around for a long time, and there are at least 3 other Linux distros for PPC that I can think of(and they're current!). I bet you could get other PPC OSes running on a Mac with little trouble(I haven't tried, so I don't know for sure).
I'm quite sure Apple hasn't abandoned PPRP/CHRP.
*PPRP stands for PowerPC Reference Platform, and was a precursor to CHRP back in the early days of Mac clones.
Oh, you mean this software?
http://osx.freshmeat.net/
True, not all of the 28,888 projects on freshmeat run on Mac OS X. But the OS X-specific section continues to grow, and many *nix applications have been ported or simply work already.
WMBC freeform/independent online radio.
My speculation: The 'optical audio' that this ad touts is an implementation of Yamaha's mLAN, a joint project between Apple and Yamaha begun in 1999.
mLAN essentially allows the transfer of all audio-related signals - be they MIDI, audio, whatever - over 1 firewire cable.
yamaha press release, mlan, 2000 [opens in new window]
Now, why would Apple release a G5-based PPC with a dedicated mLAN port? I think Apple's hiring of Doug Wyatt - the guy who invented MIDI Timecode at Opcode - as well as Apple's aquisition of eMagic - in addition to their collaboration with Yamaha on the mLAN spec - would give Apple every incentive to put an 'mLAN' port on the back of their computer, even if it is only another firewire port.
Keep in mind that OSX has MIDI capability built-in - unlike any other OS. ALso, with the addition of a simple mLAN port, Apple can now state that their PPC is music-production ready right out of the box.
Doug Wyatt hired by apple
eMagic Corporate info
You're looking at PCI-X 2.0, which will run at 266/533MHz (and higher) when it's released. The PCI-X 1.0 spec (66/133MHz) is not as insanely fast as 2.0 or Express, but it has the slight advantage of being already in use.
Since we're talking about rumours, this one here suggests that PPC 970s are cheaper for Apple to acquire than G4s by 25-35 per cent. While that in itself may have no bearing on the final price point of these new machines -- other technologies on the mobo might also have something to do with that -- it suggests, contrary to what some are expecting, that Apple might be able to at least hold prices at or near to current price points. Couple that with the fact that Apple really does want to boost sales of its Power Macs and has been somewhat more aggressive on price lately than they used to be.
That makes these machines the equivalent of a 6-10 GHz machine.
Almost. The dual 2GHz G5 would be like a 6-10 GHz Pentium 4.
Quite honestly, this would be true of any dual 2GHz UltraSPARC III, Itanium 2, or recent Alpha (if these CPUs ran that that clock rate). If you extrapolate the numbers at spec.org, the Pentium 4 looks really weak by comparison (e.g., the Alpha fp-rate numbers blew me away--it's really too bad HP is marketing the Itanic). Even the often-slammed UltraSPARC III is a fp-rate monster (it just lags in the integer stuff).
Healthcare article at Kuro5hin
None to be found anywhere.
Certainly no links to be found about Mac CAD.
What a shame.
NetInfo connection failed for server 127.0.0.1/local
The PPC 970 bus is not HyperTransport (based on released information from IBM). It can however be easily interfaced to HT via a bridge chip. I do believe that the memory bus will not be HT but more direct, the rest will utilize HT as needed... we shall see.
The show will be broadcasted at:
http://www.apple.com/quicktime/qtv/wwdc03/
>> Had I been going to bed earlier every night? Have I been sleeping later? Has Tyler been in charge longer and l
That will not be a live broadcast.
Apple will run a loop of the broadcast starting a few hours after the live presention is over.
As below, so above and beyond, I imagine drawn beyond the lines of reason. Push the envelope. Watch it bend.
Yes it is backwards compatible. It does use an improved connector, but I believe they currently ship 800/400 adapters with all of their Macs that have 800 on them. The reason to still have 400 on it is that it typically hasn't been maxxed out yet and the iPod still has a 400 connector on one end.
mbbac
But you can't build a dual-P4 at all. For that, you gotta go with a Xeon, whose price will make Apple's look quite small.
The performance will be quite different from your P4/2.8, as well. Until realworld benchmarks on real production machines come out, comparing an PPC970 Apple to an Intel or AMD machine is really quite silly. And that's not even comparing a 64-bit OS X (which by rumours may be _after_ the 64-bit hardware).
I doubt the prices will change much. Apple's prices have little to do with how much the CPU costs (and the CPUs probably cost less than the previous Moto chips). That's what the market will bear, so that's what they charge. *shrug*
Umm, have you seen the specs on a PPC970 kiddo?
The PPC970 wit its Power4 core, clocked at 1.6GHz completely trashes a 3GHz P4. Faster bus, faster integer, and completely outclasses the P4 for FPU and SIMD.
And it looks like Apple's going to ship a dual 2.0 GHz. This ain't your grandma's G4 (In fact, at the same clock, it looks like the PPC970 has a 1.5x or more advantage for integer and 2-2.5x advantage for FPU/SIMD over the G4, and the G4 is, clock-for-clock, the fastest CPU currently in the desktop and laptop market, it's only real disadvantages are low clock speed and the slow system bus, both of which are problems the PPC970 doesn't have).
Remember that Athlon is only clocked a couple of hundred MHz faster than the 970, and isn't nearly as fast, clock-for-clock.
"You've got an invalid haircut" -Warren Zevon - Life'll Kill Ya
VAIOs *are* desktop machines.
Heute die Welt, morgen das Sonnensystem!
bzzt! price/performance/productivity discussions are enless, but some of your facts are wrong.
Jun 1988: Mac OS 6.0
Oct 1990: Mac OS 6.0.7
May 1991: Mac OS 7.0
Oct 1991: Windows 3.0
Apr 1992: Windows 3.1
Aug 1992: Mac OS 7.1
June 1994: Mac OS 7.5
Aug 1995: Windows 95
Mac OS 6.x and Windows 3.1 never went head to head as shipping operating systems - Mac OS 7.0 was released nearly 6 months before even Win 3.0 . Win 95 didn't make an appearance for four and a half years after System 7.0.
*if* these machines are true, i would expect them to roughly follow the current prices. maybe the top of the line custom workhorse will jump up to a higher bracket, but in general i bet they start at the same place. i don't know about all the components, but what i have read about the PPC970 chips is that they actually cost a lot less than the G4s (as well as use less power and consume less energy). If the last statement is true, it makes you wonder how loing the G4s will be around for anything but upgrades to existing machines (seems the 970 will not work with current motherbopards).
macs *usually* fall into a pricing structure and somewhat stay the same after updates. it's not like these (if they are real) will be stacked on top of the current machine's prices and start at $4,000..... the G4 some poor citizen buys today will be dropped in price by hundreds of dollars monday if these show up. sometimes they sort of shift up and down a step according to conditions of the market. For example, for a long time there has been an emac/imac right at or around $999. I guess this pricing method is what caused their price protection deal where if you buy a machine and the price officially drops in 2 or 4 weeks (depends on who you ask?) they will refund the change.
UltarSPARC III got a high fprate by having the only compiler that was smart enough to run a loop in one of the benchmarks 'backwards', so the array striding occurred efficiently. If you look at the individual scores, you'll see one that stands out as an outlier.
Basically, US III is a fairly slow processor - despite any marketing hype from Sun.
Galileo: "The Earth revolves around the Sun!"
Score: -1 100% Flamebait
Um, Itanium systems have a maximum clock rate that's slower than what this 'leak' puts as the PPC speed. Itanium systems are quite a bit more expensive than the PPC 970 systems likely to come out of Apple.
Oh, you want to compare 32bit P4s to 64bit 970s? Why, because you're really in the market for a 32 bit processor and need to distort everything into a P4=center of the universe worldview. Grow up. Something new is coming. It will likely run rings around Itanium. It *may* run rings around the P4s for stock 32bit applications and will very likely beat it when comparing code that's similarly optimized for the two chips.
Let's see it unveiled and let's see the benchmarks then let the chip wars begin again!
For all you people who complain that macs cost too much, it's because they use quality parts, not the cheapest thing from malaysia they could find. My friend recently bought a Dell Inspirion 4100, and the screen on it, while having a large resolution, is terrible. If you go even slightly out of center with it, the colors go all wacky. But my iBook's screen is much better, even though it's 4 years older. Quality parts people, quality parts.
- Sherman
Hard to answer, it depends on what you are doing.
For integer performance the PPC 970 should be around 1.3-1.5x clock for clock against a P4. For floating point it should be around 1.4-1.6x clock for clock against a P4. For SIMD (vector calculations) it could well into 2+x clock for clock (don't have solid numbers for this). This is based on SPECint2000 & SPECfp2000 published/estimated numbers and assume linear scaling with clock speed, neither of which is true in general.
In other words a PPC 970 @ 2GHz aligns approx. with a P4 @ 3GHz for int and floating point and a P4 @ 4+GHz for SIMD.
Also the PPC 970 can move a little more data over its FSB then a P4 can at the moment (assuming the PPC 970 @ 2GHz is true). P4s @ 3GHz can move around 6.4GB/s and PPC 970 @ 2GHz can move around 7.1GB/s (for the later I am accounting for overhead, not sure if the P4 numbers are doing the same).
Anyway we will have to see how things do once they get out into the real world. At the least the new PowerMacs should be on par with P4 based systems.
Nope not PCI Express but most likely the existing PCI-X standard (133MHz/64bit).
Very unlikely. Most likely he will show off 64 bit support, PPC 970 optimizations, HyperTransport support, evolutionary extensions to the audio sub-system, evolutionary (possibly revolutionary) extensions to Quartz 2D / QuartzExtreme, and a few new things that we havn't really thought of, etc.
As the other reply to your poster noted, the maximum addressable physical memory on a machine with 32-bit virtual addresses can be >4GB, and is, in fact, >4GB on several processors, including the PowerPC's he mentioned, as well as Pentium Pro and later x86's from Intel (and probably some 32-bit x86's from AMD as well).
Only 4GB of it can be accessed at any time, however, as linear virtual addresses are 32 bits. If you're trying to use more than 4GB of physical memory, you have to map it in and out of that 4GB window. (The segmentation hardware on x86's doesn't help, as it translates 48-bit segmented addresses into 32-bit linear virtual addresses; you'd have to mark segments that don't fit into that address space "not present", and map the pages of those segments into and out of the 4GB window in response to "segment not present" faults.)
That could be done directly by privileged code, and could be done with system calls such as mmap in non-privileged code.
I got one of the first SATA drives on the market (for company testing), and it had a little Apple logo on it.
My guess is they have SATA. We do, and we're smaller than Apple.
Fellowship 9/11
yes, yes (scroll to bottom), yes-yes-yes, yes, and yes to the titles you mention by name.
Emulation options exist for every platform you mention.