Well at this pace will see if by the time Apple comes out with their 64bit chip Intel might, just might have theirs working properly too.
Sun Blasts Intel's Itanium
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
IBM Power4 and Apple
- Thread starter arn
- Start date
- Sort by reaction score
Well at this pace will see if by the time Apple comes out with their 64bit chip Intel might, just might have theirs working properly too.
Sun Blasts Intel's Itanium
I realy hope this Power4 will be the g5. Although a G5 coming from motorola wouldn't realy be a surprice.
I wonder, this power4 has some kind of altivec, but will the apps that use altivec like photoshop or final cut pro, be able to use this kind of altivec just as good as altivec itself ? Would it have the same effect as a processor with altivec ? or the same effect as a processor without altivec ?
This is realy important, if not so, apple will never use this chip because it could harm there friendship with adobe, who did everything to make photoshop optimal for altivec...
If the power4 has the same effect as altivec, then it will be amazing fast...
i want one
I wonder, this power4 has some kind of altivec, but will the apps that use altivec like photoshop or final cut pro, be able to use this kind of altivec just as good as altivec itself ? Would it have the same effect as a processor with altivec ? or the same effect as a processor without altivec ?
This is realy important, if not so, apple will never use this chip because it could harm there friendship with adobe, who did everything to make photoshop optimal for altivec...
If the power4 has the same effect as altivec, then it will be amazing fast...
i want oneActually, I've always thought this is why earlier attempts at sneaking out the 64 bit x86 stuff have flopped. They are saving viable attractive working versions for when Apple has something similar coming out, so they can knock the wind out of Apples sails by announcing around the same time.
Right now, there is no market impetus for Intel to release a 64 bit chip- the market wants the cheap stuff, and AMD isn't threatening their market enough. It serves more value as an ace in the hole to steal thunder from Apple.
Someone else said:
I call it...Mini-G....
Altivec
The new IBM PowerPC chip has a vector-processing unit that is similar to Altivec, but not the same. As such, it would be a significant departure if Apple chose to go down this route, and would probably require a lot of work on the part of software developers, who've just had to make a substantial investment in porting to OS X.
Hmm, if this does happen, it'll take some time. As SJ said, they need to complete the migration to OS X before they can consider migrating to new hardware.
Choosing IBM over Motorola could be leaping from the frying pan to the fire - IBM are not much more committed to desktop PowerPC than Moto.
If you're going to make a big hardware change, and expect your developers to support it, you'd better be sure this one is for keeps!
The new IBM PowerPC chip has a vector-processing unit that is similar to Altivec, but not the same. As such, it would be a significant departure if Apple chose to go down this route, and would probably require a lot of work on the part of software developers, who've just had to make a substantial investment in porting to OS X.
Hmm, if this does happen, it'll take some time. As SJ said, they need to complete the migration to OS X before they can consider migrating to new hardware.
Choosing IBM over Motorola could be leaping from the frying pan to the fire - IBM are not much more committed to desktop PowerPC than Moto.
If you're going to make a big hardware change, and expect your developers to support it, you'd better be sure this one is for keeps!
Actually, if you believe IBM's interview with CNet earlier this week, Cell's taped out and into production now for testing and hardware/software design purposes.
Although, admittedly, it's not likely to be using anywhere near the entire plant's potential capacity at the moment...
Problem is, IBM is just like motorola in that it's goals are beginning to unalign themselves with apple's goals. IBM makes more than half of its money by consulting with companies right now. That's why it has sold off its hard drive division and bought that consulting firm (PriceCoopersWaterhouse?). We may just see IBM sell off its microprocessor division to Samsung, Hitachi, or some other company we may not feel comfortable with as a source of our chips. *shudder*
I hope that doesn't come to pass
I hope that doesn't come to pass
Sh*t, that's the first thing I've read on this board that actually frightened me.
What then................?
What then................?
Re: Altivec
Errr, can you point to the technical specs that justify the statements above? No? Because they haven't been released?
The chip has a vector-processing unit. Period. It APPEARS that it MAY BE similar to Altivec. It might also, in fact, BE Altivec. There is no techical reason why it couldn't be. However, given everything that we KNOW, it is most likely IBM's implementation of the same instruction set that Altivec uses. That's ALL WE KNOW about the vector portions because that's all they've leaked. It might compatable with Altivec or it might not. And just because some other web site says it's not DOESN'T MEAN ANYTHING.
One other thing we do know is that this chip already has a large customer. IBM would not undertake the significant investment in the HOPE that someone would buy this chip. Someone has already placed an order. And that customer is NOT the embedded market. It's a desktop/server customer. Apple is THE ONLY current maker of PowerPC desktop computers. The only other signficant user of Power/PowerPCs is IBM. So either IBM is producing this chip for themselves and they plan on using it to enter the PowerPC desktop and low-end server market (probably with Linux), or the chip was made for Apple.
Exploring the IBM/Linux angle just doesn't make a whole lot of sense. Exploring the IBM/Mac angle makes nothing but sense.
So, I'll say it again. The chip is for Apple. Apple has already placed a large order. At the VERY outside, we'll see this in new Macs within one year. Much more probably by MWSF. Very possible by next week.
The G4 will still be around in entry-level Powerless (in comparison) macs and we'll probably see the G4 scale up to 1.6 or 1.8MHz over the next year for those machines with better bus architecture, etc., etc.
Errr, can you point to the technical specs that justify the statements above? No? Because they haven't been released?
The chip has a vector-processing unit. Period. It APPEARS that it MAY BE similar to Altivec. It might also, in fact, BE Altivec. There is no techical reason why it couldn't be. However, given everything that we KNOW, it is most likely IBM's implementation of the same instruction set that Altivec uses. That's ALL WE KNOW about the vector portions because that's all they've leaked. It might compatable with Altivec or it might not. And just because some other web site says it's not DOESN'T MEAN ANYTHING.
One other thing we do know is that this chip already has a large customer. IBM would not undertake the significant investment in the HOPE that someone would buy this chip. Someone has already placed an order. And that customer is NOT the embedded market. It's a desktop/server customer. Apple is THE ONLY current maker of PowerPC desktop computers. The only other signficant user of Power/PowerPCs is IBM. So either IBM is producing this chip for themselves and they plan on using it to enter the PowerPC desktop and low-end server market (probably with Linux), or the chip was made for Apple.
Exploring the IBM/Linux angle just doesn't make a whole lot of sense. Exploring the IBM/Mac angle makes nothing but sense.
So, I'll say it again. The chip is for Apple. Apple has already placed a large order. At the VERY outside, we'll see this in new Macs within one year. Much more probably by MWSF. Very possible by next week.
The G4 will still be around in entry-level Powerless (in comparison) macs and we'll probably see the G4 scale up to 1.6 or 1.8MHz over the next year for those machines with better bus architecture, etc., etc.
There's a long way to go from the end of the design phase to samples. And an even longer way to go from samples to shipping. From the same CNet article, "At this rate, commercial production of Cell could come as soon as the end of 2004." They're being generous with that statement.
I really doubt that IBM would let that factory sit idle for the next 2+ years. They already have plans for it. And the plans are to procduce the chip for my next Mac! Now quit trying to bring me down!
Originally posted by topicolo
Problem is, IBM is just like motorola in that it's goals are beginning to unalign themselves with apple's goals. IBM makes more than half of its money by consulting with companies right now. That's why it has sold off its hard drive division and bought that consulting firm (PriceCoopersWaterhouse?). We may just see IBM sell off its microprocessor division to Samsung, Hitachi, or some other company we may not feel comfortable with as a source of our chips. *shudder*
IBM is DEEPLY committed to the Power/PowerPC design. They invented the thing. It powers all their high-powered stuff. If IBM sold their microprocessor division, they would be handing over their crown jewels to another company. IBM would then be depending on that company to keep up development of that design for IBM's servers. NEVER GONNA HAPPEN.
Actually, I'd say they're working to hit Sony's 2005 release date for the PS3, which likely means queuing about 6 months of production to prevent the under runs they had with the PS2 which they likely won't be able to get away with the second time around as MS and Nintendo are likely to do simultaneous launches to prevent Sony getting another 30 Million console lead. If Sony plays true to form, they'll want to launch early 2005 in Japan, and late in NA to hit the Christmas season...
Sony's gonna need a pile of those chips before they go 'commercial' for hardware dev. Another pile to do software and compiler testing, since Cell's design dictates an entirely different way of programming. Then they're gonna need another pile to get dev machines in the hands of the Game companies at least a year ahead of general production so there's games on the shelves when the console ships...
Of course, all this production can be tossed between Toshiba, IBM, and I'm fairly sure Sony's got their own fab or two lying around somewhere...
Like I said, I doubt cell production is gonna be tying up the entire plants output for a while, but I doubt the plant is sitting idle...
Besides, If I wanted to depress ya, I'd point out that IBM still wants to shrink the original Power4 to make it go faster.
Or that they could aways just rent the idle portions of the plant out to one of the fabless semiconductor companies like nVidia for a profit.
Or AMD might need some extra production facilities for Clawhammer...
What I'm pretty sure of, is that we can see this chip in mass production by the end of the year. It's just a matter of taking things off the current Power4 so that the chip would be a feasible solution in a desktop computer... They don't really have to worry about scaling down the design to make it a small, low power processor for another year or more because there probably aren't any plans for notebook Power4's... That being said, Apple can have this machine out by early next year.
Or, in a really optimistic tone, I can see Jobs at a special Apple event on the 15th... "IBM's going to officially introduce a scaled down Power4 chip in October, but we're going to unofficially introduce it here, in our new PowerMac G5..."
I remember when I used to jump with joy on the news of new Apple hardware...
Or, in a really optimistic tone, I can see Jobs at a special Apple event on the 15th... "IBM's going to officially introduce a scaled down Power4 chip in October, but we're going to unofficially introduce it here, in our new PowerMac G5..."
I remember when I used to jump with joy on the news of new Apple hardware...
Like the socket design that requires 700 lbs of insertion force...
But, seriously, if we can use the Sahara G3 announcement as a guideline, IBM would probably already be in production on this chip, will have samples for early 2003, and be into mass production in April/May.
Thus, leading to a July PowerMac launch if this is where the Mac is going...
Actually, I've always thought this is why earlier attempts at sneaking out the 64 bit x86 stuff have flopped. They are saving viable attractive working versions for when Apple has something similar coming out, so they can knock the wind out of Apples sails by announcing around the same time.
Right now, there is no market impetus for Intel to release a 64 bit chip- the market wants the cheap stuff, and AMD isn't threatening their market enough. It serves more value as an ace in the hole to steal thunder from Apple.
Of course, they still have to wait on Window's to give them an Operating System. Be a good thing if Linux could step up and have a GUI to go with theirs, but still, that's probably the thought with AMD. Intel won't release one until there is a version by M$ tho.
Right now, there is no market impetus for Intel to release a 64 bit chip- the market wants the cheap stuff, and AMD isn't threatening their market enough. It serves more value as an ace in the hole to steal thunder from Apple.
Of course, they still have to wait on Window's to give them an Operating System. Be a good thing if Linux could step up and have a GUI to go with theirs, but still, that's probably the thought with AMD. Intel won't release one until there is a version by M$ tho.
It seems like this would be an opportune time, if they go to the power4, to implement a new rating convention, i.e. amd's XP ratings and maybe a possiblility of skipping the "g5" moniker to a 6 ( then it would be the power_six4 with 64bit procs!)?
Does anyone know if osX 10.2 and be recompiled (easily and quickly) to make it 64 clean? If so, then Apple won't have to wait until development on a 64b version is done before introducing a 64b proc.
Lastly, what's the chance of apple using power4's in their high end machines and offering turnkey systems for shake/tremmor, FCP to bite into Avid and Discreet profits?
Does anyone know if osX 10.2 and be recompiled (easily and quickly) to make it 64 clean? If so, then Apple won't have to wait until development on a 64b version is done before introducing a 64b proc.
Lastly, what's the chance of apple using power4's in their high end machines and offering turnkey systems for shake/tremmor, FCP to bite into Avid and Discreet profits?
Wait a minute...
"The company is in a constant tug of war with Motorola, which makes most of the PowerPC chips slotted into Macs, for Apple's business."
*Really?!?!?* Gee... I thought competition was the thing that spurred the growth of x86 performance. It seems to have had the opposite effect on IBM and Motorola.
"The company is in a constant tug of war with Motorola, which makes most of the PowerPC chips slotted into Macs, for Apple's business."
*Really?!?!?* Gee... I thought competition was the thing that spurred the growth of x86 performance. It seems to have had the opposite effect on IBM and Motorola.
Of course Mac is the only PowerPC based desktop computer. That is Apple's problem. Neither Moto nor IBM are particuarly motivated to focus on Apple's needs when developing their PowerPC lines - and it's worth noting that both Moto and IBM have abandoned PowerPC altogether at various points in the past. Remember a few years back when Moto had pulled out of G3 and we all though IBM were going to be our saviours? Then suddenly it was all talk of G4 and IBM were old news. Now suddenly we're talking IBM again.
IBM has all but given up on their desktop PC business, and they invented the damn thing. The last thing they're going to do is release a PowerPC/Linux desktop. Who the hell would want one of those when you can get a nice Intel/Linux from Dell for (presumably) a fraction of the price.
Saying that IBM would never scrap PowerPC because they invented it is like saying that Compaq would never scrap Alpha.
If Apple wants to find suppliers of competively priced and specced CPUs for desktop PCs, they'll have to go to companies that specialising in designing and producing them, and who have the economies of scale to do a good job of it - namely Intel and AMD.
IBM has all but given up on their desktop PC business, and they invented the damn thing. The last thing they're going to do is release a PowerPC/Linux desktop. Who the hell would want one of those when you can get a nice Intel/Linux from Dell for (presumably) a fraction of the price.
Saying that IBM would never scrap PowerPC because they invented it is like saying that Compaq would never scrap Alpha.
If Apple wants to find suppliers of competively priced and specced CPUs for desktop PCs, they'll have to go to companies that specialising in designing and producing them, and who have the economies of scale to do a good job of it - namely Intel and AMD.
The purpose of pointing out the facts that (a) the Mac is the only desktop PowerPC out there and (b) New IBM desktops are unlikely was to point out that Apple is far and away the mostly likely purchaser of the new chip. As pointed out before, this chip already has a customer. That customer is Apple. Do you understand now?
As far as IBM abandoning the PowerPC and Apple, you're so very far off the mark. IBM has been happy to develop and build whatever Apple asks them to. It's the Apple asking that has been in the way. Meanwhile, IBM continued to develop high-powered Power architecture chips at the same time Motorola abandoned that market in favor of the embedded market. The only reason this chip can exist at all is because of IBMs commitment to that chip.
You don't seem to know ANYTHING about IBM's business. IBM NEEDS the Power architecture for THEIR OWN machines. They will NOT sell it. They will NOT abandon it. IBM makes MUCH MUCH MUCH more than simple desktops. You do know that IBM makes computers other than desktops, don't you? That's what they do. And they make a HECK of a lot of money doing it.
Oh gosh. There are SOOOO many reasons why this WILL NEVER HAPPEN. Why am I even arguing with you? Waste of time....
May not be only for apple!
http://www.theregister.co.uk/content/4/26611.html
Tho Amiga might not ship that many computers so its profitable for IBM..That is if they ship... They could just sell there OS to run on existent PPC computers/servers
http://www.theregister.co.uk/content/4/26611.html
Tho Amiga might not ship that many computers so its profitable for IBM..That is if they ship... They could just sell there OS to run on existent PPC computers/servers
I'm not questioning the fact that IBM makes servers. I'm questioning IBMs commitent to developing PowerPC chips that meet Apple's needs. I don't believe that it is not correct to say that IBM has always made what Apple wanted. If this was the case, I think we would have seen IBM supporting Altivec a long time ago.
Whilst I recognise that Apple will not release a PC compatible version of OS X any time soon, that does not preclude the possibility of Apple migrating the Mac platform to an Intel processor. Clearly software would need to be recompiled and some software developers (Microsoft) may not be willing to do that.
All I'm saying is that if the Mac has to migrate to a new CPU - Apple might as well choose one with a future.
This is a discussion, not an arguement - I'm interested to know your views
Whilst I recognise that Apple will not release a PC compatible version of OS X any time soon, that does not preclude the possibility of Apple migrating the Mac platform to an Intel processor. Clearly software would need to be recompiled and some software developers (Microsoft) may not be willing to do that.
All I'm saying is that if the Mac has to migrate to a new CPU - Apple might as well choose one with a future.
This is a discussion, not an arguement - I'm interested to know your views