I think NMR is a little off on the date. Based on info I've heard, MWSF seems a little more likely. There have been many changes to the internal roadmap for Apple. Very early marketing materials showed that Gigawire was supposed to make a debut in the current machines. But that changed after a couple of months. Due to circumstantial info, I will speculate that the POWERmac project is going along much faster than previously estimated.
the combination of IBM and Apple to make a 64bit chip for desktops et al makes sense when you put the "IBM loves Linux" equation into it. Also, if its true that IBM is putting into GCC as well... well, looks like one hell of a nice partnership seeing that IBM would provide Apple chips, and also ITSELF (and anyother Linux/Unix provider).
This makes more sense then the current situation wehere basically Apple is the only real client for PPC chips (aside form all the embedded stuff).
This is all coming together folks.... but the skeptic in me says that by next month, what seemed like the umpteenth "GREAT IDEA" is going to fizzle into nothingness and in 3 years we will still be using G4s. Blah.
1.) Apple becomes unable to keep OSX on PPC-Linux emulators at bay. Linux-PPC becomes a meaningful competitor to Wintel and cheap boxes are available. We all buy cheap generic boxes and run OSX on them. Steve Jobs has a coronary, partially recovers, and is consigned to a mental asylum. WOO-HOO!
2.)IBM is interested in going after M$ and not relying on M$ technologies. They can develop Linux to run on PPC's (which they control) and possibly make a better flavor of Linux than any of the "run-it-on-cheap-X86-boxen" versions out there, but it will take time. Maybe IBM can transition a major portion of it's business clients to PPC-Linux. They have Lotus and could whip up a 'standard enough' (read M$ compatible) suite. In the meantime Apple at least gets timely CPU updates.
3.) IBM buys Apple. Rather than go through the trouble of developing linux, IBM buys a major stake in Apple. They combine brands. IBM markets OSX to business and corporate customers and Apple markets to home users and creative pros. IF (big IF, but possible) IBM could transition all of it's clients to IBM-PPC-linux/OSX machines, Mac could double or triple it's market share (into the 10% region) virtually overnight. Of course, IBM has not been very good at pushing its proprietary standards to the masses -- they're part of the reason M$ is now the goliath it's become.
hmmm... it all seems unlikely.
Though, it'll be a great day when Steve is no longer able to physically close off the platform. hehehe...
NewWorld Macs are pretty much generic OpenFirmware PowerPCs.
There are tonnes of Linux distros for Macs, and I'm not aware of any technical impediment to writing generic PowerPC motherboard drivers for Mac OS X.
What would stop it is Apple lawsuites agains ppl selling Mac OS X copies which have the ability to run on non-Macs. Apple would probably let the geeky mod their personal copies of X.
MOL is just too geeky for the masses. The Mac is about plug it in, turn it on, start working. MOL means building a box, installing an OS, blah blah blah...
I guess the only possible solution is a Linux distro which boots then launches Mac OS X inside MOL transparently. But then it would only be as fast as a Mac which costs less, so who would care?
<strong>1.) Apple becomes unable to keep OSX on PPC-Linux emulators at bay. Linux-PPC becomes a meaningful competitor to Wintel and cheap boxes are available. We all buy cheap generic boxes and run OSX on them.</strong><hr></blockquote>
Not going to happen. You're thinking of the Lindows scenario which is possible because MS does not conrol the hardware. Apple does. It would be a trivial thing to ensure an OS release that breaks the most recent Linux-PPC software for generic PPC mobos. Something as simple as requiring a proprietary ROM in order to boot, for instance. (Sound familiar?)
[quote]<strong>2.)IBM is interested in going after M$ and not relying on M$ technologies. They can develop Linux to run on PPC's (which they control) and possibly make a better flavor of Linux than any of the "run-it-on-cheap-X86-boxen" versions out there, but it will take time.</strong><hr></blockquote>
Heh. How about a PPC version of OS2? (Call it OS3?)
[quote]<strong>3.) IBM buys Apple. Rather than go through the trouble of developing linux, IBM buys a major stake in Apple. They combine brands. IBM markets OSX to business and corporate customers and Apple markets to home users and creative pros.</strong><hr></blockquote>
Why would IBM spend 5-6 Billion to buy a Unix variant OS when they already have one? That's a lot of cash to buy a company with a minority market share position.
[quote]<strong>Though, it'll be a great day when Steve is no longer able to physically close off the platform. hehehe...</strong><hr></blockquote>
Never going to happen. Even if Apple migrates to x86 CPUs, they'll still make their own mobos and chip sets. No one will be able to run Mac OS X on anything other than Apple hardware regardless of the CPU's ISA. Discard all notions of clones. Not going to happen.
[ 09-16-2002: Message edited by: Tomb of the Unknown ]</p>
Yeah, like I said, I don't think any of it would happen, but I do think MOL will make it virtually impossible for Apple to keep OSX off other machines. Could they sue MOL? Can Intel sue over VPC? Maybe. However, I just don't see the huge performance hit of running OSX in MOL? It's all PPC, it's not like you're emulating a different chip architecture. The only thing thats missing is cheap PPC based desktops. If we get 'em, someone will make it work, and make it work well.
<strong>Yeah, like I said, I don't think any of it would happen, but I do think MOL will make it virtually impossible for Apple to keep OSX off other machines. Could they sue MOL? Can Intel sue over VPC? Maybe. However, I just don't see the huge performance hit of running OSX in MOL? It's all PPC, it's not like you're emulating a different chip architecture. The only thing thats missing is cheap PPC based desktops. If we get 'em, someone will make it work, and make it work well.</strong><hr></blockquote>
Why would I waste my time and money on MOL when I can run OSX natively.
Despite the ravings and rantings of us here on this board Apple DOES have a complete range of PPC hardware. I don't expect for them to give their "Crown Jewel" the Powermac away for next to nothing.
MacOS X has made running Linux a moot point.
The eMac has made wishing for Intel like prices a moot point(Anyone who doesn't have PC envy naturally).
With the possible entry of the The G5(let's get this back on Topic) I see Apple being able to use higher speed G4's in the consumer offerings like iMac, eMac Powerbook. The problem with Apples lineup has always been not enough range in proc speeds to offer competitive speeds at the lowend.
"With the possible entry of the The G5(let's get this back on Topic) I see Apple being able to use higher speed G4's in the consumer offerings like iMac, eMac Powerbook. The problem with Apples lineup has always been not enough range in proc speeds to offer competitive speeds at the lowend."
...and that looks like it's aobut to change relatively soon...
Hmurchinson, as usual, your blind Apple apology lets you gleefully miss the point. 799? Really, for a 2 year out of date unexpandable fuzzy 15" CRT? Please.
IF affordable PPC linux boards were available, the whole point would be to get a fast CPU/mobo and load it with my own RAM, HDD and Video cards in a case of my choice. Then run Linux on it, and use that machine to boot up OSX in a more advanced version of MOL (or something like it). IF PPC-linux gained any kind of wide-spread adoption, this would be possible, and I guarantee that it would be doable at a much lower price than equivalent speed Apple machines. Apple knows that it would be virtually impossible to lock generic machines out of OSX if they switched to X86. Box makers aren't interested in PPC, but, IF they were, Apple would have a problem and they know it.
If an alternative PPC platform became popular, Apple would have all the same troubles that a potential switch to X86 would bring.
I'm not an Apple Apologist..I just feel that Apple get's misrepresented too much.
I'm not going to jump through flaming hoops just to run OSX and save a few bucks on Hardware.
You know most people time is precious. Computer Enthusiasts have time to build and tinker with computers but a vast majority need a good OOBE. MoL and a DIY box isn't gonna fly with them.
If Apple really wanted to lock OSX away from other PPC mobo's surely it is possible by adding something to there boards that can only be bypassed by reverse engineering, which would spark a legal row...
Surely this is a situation thats not disimilar to the problems that BeOS faced when apple refused access to the code to let BeOS 5 run on the G3...
I might be wrong (probably am) but just wanted to chip in.
Back on the G5/IBM track... It's good to hear that so many rumors that have floated around are tying together a bit.
I believe that the change to OSX only boot is more of a marketing ploy and an tip of the hat to MS to encourage adoption, hence the InDesign bundle.
I have a question though...is there any indication of what speeds the GPUL chip is being pitched at and what it might scale too?
<strong>I have a question though...is there any indication of what speeds the GPUL chip is being pitched at and what it might scale too?
</strong><hr></blockquote>
There are no details yet. If the processor is based on the POWER4 core then it could have 16 stages. But if the processor is more closely related to Rivina then it may have as few as 6 pipeline stages. i think the design will fall somewhere in between and will compete nicely with the P4 and Athlon in terms of MHz... but real performance will make this processor really shine. One of the few details we do know is that it's a 8 way super-scalar processor; 8 instructions dispatched be clock cycle. but as mentioned earlier the real number may be closer to 5 or 6 actually completed instructions, but even so this puts it well ahead of any other workstation processor.
I would estimate the speed range to be between 1.3 and 2.0GHz on a 130nm process.
"x86 is the least interesting rumor, though, as it really means nothing than recompiling Mac OS X and deploying on existing hardware."
That must be one of the most persistent myths on the Apple boards...
And in most cases a 32-bit application will run SLOWER on a 64-bit box.
As for rumors, the observation that the mood on these boards change abruptly with each new batch of rumors, is very succinctly put. The only two reasons reason to read the declarations of "sources" or "friends of sources" is to get entertained, and to get enlightened by the knowledgeable and well-considered postings of people like Programmer (and several others whose nicks I don't remember at this moment)...
oKAY - Moki, you're BACK! :-) Good to "see" you. Guess you just missed an elevation to sainthood there...
Outsider. Can the current G4 Dispatch AND Complete 4 intructions? How much do you estimate the additional 8/5 Superscalar features of GP-UL will have on this procs performance?
Thanks for those links. Sounds VERY interesting. Apple will have to work hard to keep this in the bag if they are going to use this IBM proc.
Comments
What exactly is Gigawire?
for Acronyms: GIGA- = 10 to 9 (huh?)
It also means, as we know, one billion.
It comes from the Greek which means GIANT
Does this help in anyway? I have no clue and I doubt it...
This makes more sense then the current situation wehere basically Apple is the only real client for PPC chips (aside form all the embedded stuff).
This is all coming together folks.... but the skeptic in me says that by next month, what seemed like the umpteenth "GREAT IDEA" is going to fizzle into nothingness and in 3 years we will still be using G4s. Blah.
Like an AMD PowerPC?
Oh yeah, AMD is too busy developing and producing great x86 CPUs.
Like a desktop Motorola G5?
Oh yeah, Motorola seems to have lost all interest in Apple (and Apple in Motorola).
Like a switch to x86?
Oh yeah, too soon after the X migration.
The latest great idea: The GPUL.
Now, we know that IBM is actually making the first desktop PowerPC in many years.
Why the f*** won't Apple be using it? (Given what I outlined above).
Barto
1.) Apple becomes unable to keep OSX on PPC-Linux emulators at bay. Linux-PPC becomes a meaningful competitor to Wintel and cheap boxes are available. We all buy cheap generic boxes and run OSX on them. Steve Jobs has a coronary, partially recovers, and is consigned to a mental asylum. WOO-HOO!
2.)IBM is interested in going after M$ and not relying on M$ technologies. They can develop Linux to run on PPC's (which they control) and possibly make a better flavor of Linux than any of the "run-it-on-cheap-X86-boxen" versions out there, but it will take time. Maybe IBM can transition a major portion of it's business clients to PPC-Linux. They have Lotus and could whip up a 'standard enough' (read M$ compatible) suite. In the meantime Apple at least gets timely CPU updates.
3.) IBM buys Apple. Rather than go through the trouble of developing linux, IBM buys a major stake in Apple. They combine brands. IBM markets OSX to business and corporate customers and Apple markets to home users and creative pros. IF (big IF, but possible) IBM could transition all of it's clients to IBM-PPC-linux/OSX machines, Mac could double or triple it's market share (into the 10% region) virtually overnight. Of course, IBM has not been very good at pushing its proprietary standards to the masses -- they're part of the reason M$ is now the goliath it's become.
hmmm... it all seems unlikely.
Though, it'll be a great day when Steve is no longer able to physically close off the platform. hehehe...
<strong>The typical assumption is that if 32bit apps are better than 16bit apps. Then 64bit apps are better than 32bit.
Well I'd suppose if you had extraordinary needs.</strong><hr></blockquote>
Or if you are running high end database software that requires you to cache large amounts of data in memory.
I had wondered why Sybase and Oracle were developing more than client applications for Mac OS X. I think I know now.
Wonder how long it will be before we see SQL Server for Mac OS X?
NewWorld Macs are pretty much generic OpenFirmware PowerPCs.
There are tonnes of Linux distros for Macs, and I'm not aware of any technical impediment to writing generic PowerPC motherboard drivers for Mac OS X.
What would stop it is Apple lawsuites agains ppl selling Mac OS X copies which have the ability to run on non-Macs. Apple would probably let the geeky mod their personal copies of X.
MOL is just too geeky for the masses. The Mac is about plug it in, turn it on, start working. MOL means building a box, installing an OS, blah blah blah...
I guess the only possible solution is a Linux distro which boots then launches Mac OS X inside MOL transparently. But then it would only be as fast as a Mac which costs less, so who would care?
Barto
<strong>1.) Apple becomes unable to keep OSX on PPC-Linux emulators at bay. Linux-PPC becomes a meaningful competitor to Wintel and cheap boxes are available. We all buy cheap generic boxes and run OSX on them.</strong><hr></blockquote>
Not going to happen. You're thinking of the Lindows scenario which is possible because MS does not conrol the hardware. Apple does. It would be a trivial thing to ensure an OS release that breaks the most recent Linux-PPC software for generic PPC mobos. Something as simple as requiring a proprietary ROM in order to boot, for instance. (Sound familiar?)
[quote]<strong>2.)IBM is interested in going after M$ and not relying on M$ technologies. They can develop Linux to run on PPC's (which they control) and possibly make a better flavor of Linux than any of the "run-it-on-cheap-X86-boxen" versions out there, but it will take time.</strong><hr></blockquote>
Heh. How about a PPC version of OS2? (Call it OS3?)
[quote]<strong>3.) IBM buys Apple. Rather than go through the trouble of developing linux, IBM buys a major stake in Apple. They combine brands. IBM markets OSX to business and corporate customers and Apple markets to home users and creative pros.</strong><hr></blockquote>
Why would IBM spend 5-6 Billion to buy a Unix variant OS when they already have one? That's a lot of cash to buy a company with a minority market share position.
[quote]<strong>Though, it'll be a great day when Steve is no longer able to physically close off the platform. hehehe...</strong><hr></blockquote>
Never going to happen. Even if Apple migrates to x86 CPUs, they'll still make their own mobos and chip sets. No one will be able to run Mac OS X on anything other than Apple hardware regardless of the CPU's ISA. Discard all notions of clones. Not going to happen.
[ 09-16-2002: Message edited by: Tomb of the Unknown ]</p>
IBM integrates the software and hardware into mainframes.
Apple integrates the software and hardware into Macs.
IBM switches its business customers over to Macs and Mac OS X IBMs.
This is just an example, but Apple is far, FAR more than a simple *nix vendor.
Barto
<strong>Yeah, like I said, I don't think any of it would happen, but I do think MOL will make it virtually impossible for Apple to keep OSX off other machines. Could they sue MOL? Can Intel sue over VPC? Maybe. However, I just don't see the huge performance hit of running OSX in MOL? It's all PPC, it's not like you're emulating a different chip architecture. The only thing thats missing is cheap PPC based desktops. If we get 'em, someone will make it work, and make it work well.</strong><hr></blockquote>
Cheap PPC Desktops
<a href="http://store.apple.com/1-800-MY-APPLE/WebObjects/AppleStore.woa/241/wo/m9AHP0D1GfK4FSQMlj1/0.3.0.3.34.39.3.3.1.1.0?105,45"; target="_blank">Take your pick </a>
$799 Entry iMac
Why would I waste my time and money on MOL when I can run OSX natively.
Despite the ravings and rantings of us here on this board Apple DOES have a complete range of PPC hardware. I don't expect for them to give their "Crown Jewel" the Powermac away for next to nothing.
MacOS X has made running Linux a moot point.
The eMac has made wishing for Intel like prices a moot point(Anyone who doesn't have PC envy naturally).
With the possible entry of the The G5(let's get this back on Topic) I see Apple being able to use higher speed G4's in the consumer offerings like iMac, eMac Powerbook. The problem with Apples lineup has always been not enough range in proc speeds to offer competitive speeds at the lowend.
...and that looks like it's aobut to change relatively soon...
Lemon Bon Bon
PS. Good thread folks.
IF affordable PPC linux boards were available, the whole point would be to get a fast CPU/mobo and load it with my own RAM, HDD and Video cards in a case of my choice. Then run Linux on it, and use that machine to boot up OSX in a more advanced version of MOL (or something like it). IF PPC-linux gained any kind of wide-spread adoption, this would be possible, and I guarantee that it would be doable at a much lower price than equivalent speed Apple machines. Apple knows that it would be virtually impossible to lock generic machines out of OSX if they switched to X86. Box makers aren't interested in PPC, but, IF they were, Apple would have a problem and they know it.
If an alternative PPC platform became popular, Apple would have all the same troubles that a potential switch to X86 would bring.
I'm not going to jump through flaming hoops just to run OSX and save a few bucks on Hardware.
You know most people time is precious. Computer Enthusiasts have time to build and tinker with computers but a vast majority need a good OOBE. MoL and a DIY box isn't gonna fly with them.
Just thinking about the PPC/Linux thing...
If Apple really wanted to lock OSX away from other PPC mobo's surely it is possible by adding something to there boards that can only be bypassed by reverse engineering, which would spark a legal row...
Surely this is a situation thats not disimilar to the problems that BeOS faced when apple refused access to the code to let BeOS 5 run on the G3...
I might be wrong (probably am) but just wanted to chip in.
Back on the G5/IBM track... It's good to hear that so many rumors that have floated around are tying together a bit.
I believe that the change to OSX only boot is more of a marketing ploy and an tip of the hat to MS to encourage adoption, hence the InDesign bundle.
I have a question though...is there any indication of what speeds the GPUL chip is being pitched at and what it might scale too?
Cheers!
<strong>I have a question though...is there any indication of what speeds the GPUL chip is being pitched at and what it might scale too?
</strong><hr></blockquote>
There are no details yet. If the processor is based on the POWER4 core then it could have 16 stages. But if the processor is more closely related to Rivina then it may have as few as 6 pipeline stages. i think the design will fall somewhere in between and will compete nicely with the P4 and Athlon in terms of MHz... but real performance will make this processor really shine. One of the few details we do know is that it's a 8 way super-scalar processor; 8 instructions dispatched be clock cycle. but as mentioned earlier the real number may be closer to 5 or 6 actually completed instructions, but even so this puts it well ahead of any other workstation processor.
I would estimate the speed range to be between 1.3 and 2.0GHz on a 130nm process.
"x86 is the least interesting rumor, though, as it really means nothing than recompiling Mac OS X and deploying on existing hardware."
That must be one of the most persistent myths on the Apple boards...
And in most cases a 32-bit application will run SLOWER on a 64-bit box.
As for rumors, the observation that the mood on these boards change abruptly with each new batch of rumors, is very succinctly put. The only two reasons reason to read the declarations of "sources" or "friends of sources" is to get entertained, and to get enlightened by the knowledgeable and well-considered postings of people like Programmer (and several others whose nicks I don't remember at this moment)...
oKAY - Moki, you're BACK! :-) Good to "see" you. Guess you just missed an elevation to sainthood there...
engpjp
Look at the very last entry on the whole page.
<a href="http://www.midrangeserver.com/tfh/tfh030402-story02.html"; target="_blank">http://www.midrangeserver.com/tfh/tfh030402-story02.html</a>;
Last paragraph states that Power4-II servers should be out sometime in October of 2002...
Thanks for those links. Sounds VERY interesting. Apple will have to work hard to keep this in the bag if they are going to use this IBM proc.