<ol type="1">[*]Won't be cheap. They'll be priced like Xeon's (I *hope* not Itaniums) and workstattions based on them will be priced accordingly.[*]It will be another processor that Apple uses (if it does) in which Apple is not really part of the target market.[/list=a]
In other words, if Apple can manage the memory controller design (no sure thing), they may have the next $6,000 XServe.
I have dreams of a nice 970, but I strongly suggest people start to manage their expectations. Most of what IBM has stated about the chip indicate that this will not be an ideal chip for Apple to use. The Power4 is a nice chip too, but it doesn't mean that Apple can use it.
The only bright spot is they did add Altivec, which means that they hope Apple will use it *somewhere*.</strong><hr></blockquote>
[quote]Originally posted by Barto:
<strong>
Take my word for it. THIS CHIP WAS FREAKIN' BUILT FOR APPLE.
Barto
PS Expect Mac OS X Server to be licensed to IBM to compete with Xeons.</strong><hr></blockquote>
I expect the truth is somewhere in the middle. I'd guess the 970 has been built with the expectation that Apple will use it (with Apple having had some input into its design), although Apple will only be one of a number of customers. It may not be the ideal chip for Apple, but it's better than any other alternative currently available. I expect IBM are subject to a non disclosure agreement, preventing Apple from being mentioned as part of its target market.
I do think peoples expectations are a little high though. I can see it being introduced into the XServe and PowerMac upon its release, but only as a single processor in the latter (except perhaps in the high end model). I think people expecting it to appear in the PowerBook and especially the iMac anytime in the next couple of years may be disappointed. Although technically possible, I expect Apple will stick with improvements to the G4.
The PowerPC 970 is NOT a super-performance single chip design. It's designed to be a middle of the road performer.
However, it's designed to be used in symetric multiprocessing.
Unlike the G4, having dual CPUs will allow Apple to match, and beat, the performance of the latest Intel-based (maybe AMD, but they seem pretty hopeless ATM) computers.
It will probably be cheap enough for Apple to use Dual 1.2GHz, Dual 1.5GHz and Dual 1.8GHz across the Power Mac (whatever it looks like) line.
It's pretty much the ideal chip for Apple. It should be cool enough for the iMac in single proc, and definatly be cool enough for the iMac and PowerBook when it goes 90nm.
Velocity Engine, 64-bit PPC and 32-bit PPC (Compatible).
5 dispatches per cycle, deep pipeline for lotsa MHz (Fast).
6.4GB/s bandwidth per processor (Ideal for scaling performance by adding more CPUs).
Don't expect IBM to license OSX for server duty, IBM has a much larger investment in Linux, and it is much more capable as a server product right now.
However, I think Barto is correct about the 970 being a mid-performance low-cost chip. It is not just for Apple, it is designed with low-power requirements in mind, it is the first of a series, and its going everywhere that the G4 goes--embedded, Apple, blade servers etc. Just because it has decent performance should not mean it will be expensive. After all, this is exactly where Moto should have been in the performance spectrum sometime last year.
Thank you for agreeing with me that it will be a low-cost chip. However, it will be more expensive to implement than the G4. Embedded customers will not buy it because they don't need anything faster than the G4.
The PowerPC 970 is a "desktop and entry-level server chip".
The G4 is a "high-end embedded chip".
Linux is in some areas more mature than Mac OS X Server. However, it just doesn't have the user interface. IBM is about options, and Mac OS X Server will probably be an option for server geeks who want a great user interface.
<strong> The PowerPC 970 is NOT a super-performance single chip design. It's designed to be a middle of the road performer.
so when do we get a super-performance chip thaat's deisgned to be a killer performer? or are we forever doomed to make excuses on apple's behalf? </strong><hr></blockquote>
I think you misunderstood the comment. He means it's not a super-performace chip compared to say, it's daddy, the Power 4.
PPC970 will be a great kick butt chip. It is not going to be a Super chip that will kick intels butt all around. But for the most part it will be much more efficient and will compete at similar speeds. It will be the chip that brings apple back to the top. But its not going to crush intel into dirt...except he PC laptops which it will.
I think the comment from IBM means that IBM wants to use the 970 to compete against the Xeon. Apple can use it to compete against whatever they want, all IBM cares about that is the money they get from Apple.
Think about it -- what processors does IBM currently use the compete against the Xeon? The PowerPC 604e & POWER3. This is simply not competitive anymore, so the 970 is going to change that.
The 970 is 0.13 micron and 50 million transistors. That means it is going to be cheap compared to the Intel P4 and Xeon, which are bigger and hotter chips. Its bus is fast, but not so much faster than HT, RIO, and the new P4 FSB. The 970 is much more superscalar, much longer pipelines, much higher clockrate, and a real out-of-order execution processor. It isn't magic, but it is the latest in processor design applied to a desktop level PowerPC. The reason to get excited about this is not that it is some wonderous acheivement, but instead that it is the first new PowerPC core designed for the desktop since the 604 back in 1995. Until now we've been hanging on using essentially the same core modernized w/ AltiVec added. It also means that IBM is putting money back into PowerPC development.
No need to spread FUD here, and no cause to get excessively excited about how PowerPC is going to stomp all over Intel's latest offerings. The fact is that they are all drawing on roughly the same level of technology available and they're all going to be in roughly the same performance ballpark. Some design decisions from each camp will be more appropriate and appealing to some customers, others won't be. Apple will continue to do its best to make the most of what it is available to it, and the number of jokes about how lame Mac performance is will diminish dramatically. Steve will probably start doing more bake-offs again. Hopefully not too many more as I think they get rather tiresome after a while. Using apps other than Photoshop would be nice too.
<strong>It would be quite cool if Apple/IBM aided Connectix im making Virtual PC faster than IA64 at running x86 32 bit code. </strong><hr></blockquote>
Now that's funny. Boy would that burn Microsoft's butt. <img src="graemlins/lol.gif" border="0" alt="[Laughing]" />
<strong> The PowerPC 970 is NOT a super-performance single chip design. It's designed to be a middle of the road performer.
so when do we get a super-performance chip thaat's deisgned to be a killer performer? or are we forever doomed to make excuses on apple's behalf? </strong><hr></blockquote>
First, as others have said, "middle of the road" is in comparison to processors that destroy the P4 utterly.
Second, I'm confident that once you see what the 970 and its successors can do with SMP, you'll lose the frown. It won't be much like the G4. If VSP comes to Mac OS X via IBM, then it will truly be nothing like the current situation, because single-threaded apps will be able to use multiple processors. The question then would be efficiency.
I couldn't help laughing at the idea that Apple could get IA-32 running faster on their hardware than on IA-64. That's almost worth doing just to razz Wintel. They might decline, but there's always IBM and WINE. WINE Is Not an Emulator, but if you couple it with one...
<strong>Think about it -- what processors does IBM currently use the compete against the Xeon? The PowerPC 604e & POWER3. This is simply not competitive anymore, so the 970 is going to change that.</strong><hr></blockquote>
I don't think pSeries tries to compete with Xeons any more. If you want low-end, that's what xSeries is for.
I dont think there is anything to speculate about. What happens happens and Apple already knows what processors they will be using for the next year or so.
I can see it being introduced into the XServe and PowerMac upon its release, but only as a single processor in the latter (except perhaps in the high end model). </strong><hr></blockquote>
I tend to disagree with people that Apple would intro the 970 in the Xserve because of the Xserve market. The Xserve is Apple's attempt to get into server rooms. This is an environment that shares no relation to desktop and reliability/uptime is crucial. I would think that such a shift in processors would be allowed to mature a few months in the desktop market before selling an Xserve with it. Apple is trying to win new enterprise business and if they ship a 1st gen server without the bugs worked out, they'll shoot themselves in the foot because enterprise would never come back. This is why the Xserve is still at a slower processor speed than the desktop (since Apple is overclocking the desktop machines).
All that being said, that's just my opinion, and Apple could very well do the opposite; however, I believe that you have to be very humble and conservative competing in the server world (unlike NT server).
I tend to disagree with people that Apple would intro the 970 in the Xserve because of the Xserve market. The Xserve is Apple's attempt to get into server rooms. This is an environment that shares no relation to desktop and reliability/uptime is crucial. I would think that such a shift in processors would be allowed to mature a few months in the desktop market before selling an Xserve with it. Apple is trying to win new enterprise business and if they ship a 1st gen server without the bugs worked out, they'll shoot themselves in the foot because enterprise would never come back. This is why the Xserve is still at a slower processor speed than the desktop (since Apple is overclocking the desktop machines).
All that being said, that's just my opinion, and Apple could very well do the opposite; however, I believe that you have to be very humble and conservative competing in the server world (unlike NT server).</strong><hr></blockquote>
I agree, but for a different reason. If Apple introduces the PPC970 on the Xserve before the PowerMac, what happens to PowerMac sales in the interim? No one can be certain, but I would bet that it would not be pretty.
I think in 2001 in December Macosrumors ran a few crazy reports where they claimed the Moto G5 chip would make virtual pc run faster then normal Windows on a Dell.
Basically this will never happen. Maybe with a 5 gzh 970 but at that time Intel will have a faster chip also.
I think in 2001 in December Macosrumors ran a few crazy reports where they claimed the Moto G5 chip would make virtual pc run faster then normal Windows on a Dell.
Basically this will never happen. Maybe with a 5 gzh 970 but at that time Intel will have a faster chip also.
Comments
Slightly higher priced Powermacs? Who cares.
They'll prob' go in Powermacs.
Denial, Tom. Nothing like it, eh?
Lemon Bon BOn
<strong>
Most of what IBM has stated about the chip indicate that this will not be an ideal chip for Apple to use.</strong><hr></blockquote>
Take my word for it. THIS CHIP WAS FREAKIN' BUILT FOR APPLE.
Barto
PS Expect Mac OS X Server to be licensed to IBM to compete with Xeons.
<strong>The part I find scary is
What does this say to me? These chips
<ol type="1">[*]Won't be cheap. They'll be priced like Xeon's (I *hope* not Itaniums) and workstattions based on them will be priced accordingly.[*]It will be another processor that Apple uses (if it does) in which Apple is not really part of the target market.[/list=a]
In other words, if Apple can manage the memory controller design (no sure thing), they may have the next $6,000 XServe.
I have dreams of a nice 970, but I strongly suggest people start to manage their expectations. Most of what IBM has stated about the chip indicate that this will not be an ideal chip for Apple to use. The Power4 is a nice chip too, but it doesn't mean that Apple can use it.
The only bright spot is they did add Altivec, which means that they hope Apple will use it *somewhere*.</strong><hr></blockquote>
[quote]Originally posted by Barto:
<strong>
Take my word for it. THIS CHIP WAS FREAKIN' BUILT FOR APPLE.
Barto
PS Expect Mac OS X Server to be licensed to IBM to compete with Xeons.</strong><hr></blockquote>
I expect the truth is somewhere in the middle. I'd guess the 970 has been built with the expectation that Apple will use it (with Apple having had some input into its design), although Apple will only be one of a number of customers. It may not be the ideal chip for Apple, but it's better than any other alternative currently available. I expect IBM are subject to a non disclosure agreement, preventing Apple from being mentioned as part of its target market.
I do think peoples expectations are a little high though. I can see it being introduced into the XServe and PowerMac upon its release, but only as a single processor in the latter (except perhaps in the high end model). I think people expecting it to appear in the PowerBook and especially the iMac anytime in the next couple of years may be disappointed. Although technically possible, I expect Apple will stick with improvements to the G4.
[ 02-01-2003: Message edited by: RodUK ]</p>
However, it's designed to be used in symetric multiprocessing.
Unlike the G4, having dual CPUs will allow Apple to match, and beat, the performance of the latest Intel-based (maybe AMD, but they seem pretty hopeless ATM) computers.
It will probably be cheap enough for Apple to use Dual 1.2GHz, Dual 1.5GHz and Dual 1.8GHz across the Power Mac (whatever it looks like) line.
It's pretty much the ideal chip for Apple. It should be cool enough for the iMac in single proc, and definatly be cool enough for the iMac and PowerBook when it goes 90nm.
Velocity Engine, 64-bit PPC and 32-bit PPC (Compatible).
5 dispatches per cycle, deep pipeline for lotsa MHz (Fast).
6.4GB/s bandwidth per processor (Ideal for scaling performance by adding more CPUs).
It seems pretty ideal to me.
Barto
However, I think Barto is correct about the 970 being a mid-performance low-cost chip. It is not just for Apple, it is designed with low-power requirements in mind, it is the first of a series, and its going everywhere that the G4 goes--embedded, Apple, blade servers etc. Just because it has decent performance should not mean it will be expensive. After all, this is exactly where Moto should have been in the performance spectrum sometime last year.
The PowerPC 970 is a "desktop and entry-level server chip".
The G4 is a "high-end embedded chip".
Linux is in some areas more mature than Mac OS X Server. However, it just doesn't have the user interface. IBM is about options, and Mac OS X Server will probably be an option for server geeks who want a great user interface.
Barto
so when do we get a super-performance chip thaat's deisgned to be a killer performer? or are we forever doomed to make excuses on apple's behalf?
<strong> The PowerPC 970 is NOT a super-performance single chip design. It's designed to be a middle of the road performer.
so when do we get a super-performance chip thaat's deisgned to be a killer performer? or are we forever doomed to make excuses on apple's behalf?
I think you misunderstood the comment. He means it's not a super-performace chip compared to say, it's daddy, the Power 4.
Think about it -- what processors does IBM currently use the compete against the Xeon? The PowerPC 604e & POWER3. This is simply not competitive anymore, so the 970 is going to change that.
The 970 is 0.13 micron and 50 million transistors. That means it is going to be cheap compared to the Intel P4 and Xeon, which are bigger and hotter chips. Its bus is fast, but not so much faster than HT, RIO, and the new P4 FSB. The 970 is much more superscalar, much longer pipelines, much higher clockrate, and a real out-of-order execution processor. It isn't magic, but it is the latest in processor design applied to a desktop level PowerPC. The reason to get excited about this is not that it is some wonderous acheivement, but instead that it is the first new PowerPC core designed for the desktop since the 604 back in 1995. Until now we've been hanging on using essentially the same core modernized w/ AltiVec added. It also means that IBM is putting money back into PowerPC development.
No need to spread FUD here, and no cause to get excessively excited about how PowerPC is going to stomp all over Intel's latest offerings. The fact is that they are all drawing on roughly the same level of technology available and they're all going to be in roughly the same performance ballpark. Some design decisions from each camp will be more appropriate and appealing to some customers, others won't be. Apple will continue to do its best to make the most of what it is available to it, and the number of jokes about how lame Mac performance is will diminish dramatically. Steve will probably start doing more bake-offs again. Hopefully not too many more as I think they get rather tiresome after a while. Using apps other than Photoshop would be nice too.
[ 02-01-2003: Message edited by: Programmer ]</p>
<strong>It would be quite cool if Apple/IBM aided Connectix im making Virtual PC faster than IA64 at running x86 32 bit code.
Now that's funny. Boy would that burn Microsoft's butt. <img src="graemlins/lol.gif" border="0" alt="[Laughing]" />
Just wondering if it's possible though? Marklar??
[ 02-01-2003: Message edited by: rickag ]</p>
<strong> The PowerPC 970 is NOT a super-performance single chip design. It's designed to be a middle of the road performer.
so when do we get a super-performance chip thaat's deisgned to be a killer performer? or are we forever doomed to make excuses on apple's behalf?
First, as others have said, "middle of the road" is in comparison to processors that destroy the P4 utterly.
Second, I'm confident that once you see what the 970 and its successors can do with SMP, you'll lose the frown. It won't be much like the G4. If VSP comes to Mac OS X via IBM, then it will truly be nothing like the current situation, because single-threaded apps will be able to use multiple processors. The question then would be efficiency.
I couldn't help laughing at the idea that Apple could get IA-32 running faster on their hardware than on IA-64. That's almost worth doing just to razz Wintel.
<strong>Think about it -- what processors does IBM currently use the compete against the Xeon? The PowerPC 604e & POWER3. This is simply not competitive anymore, so the 970 is going to change that.</strong><hr></blockquote>
I don't think pSeries tries to compete with Xeons any more. If you want low-end, that's what xSeries is for.
<strong>
I can see it being introduced into the XServe and PowerMac upon its release, but only as a single processor in the latter (except perhaps in the high end model). </strong><hr></blockquote>
I tend to disagree with people that Apple would intro the 970 in the Xserve because of the Xserve market. The Xserve is Apple's attempt to get into server rooms. This is an environment that shares no relation to desktop and reliability/uptime is crucial. I would think that such a shift in processors would be allowed to mature a few months in the desktop market before selling an Xserve with it. Apple is trying to win new enterprise business and if they ship a 1st gen server without the bugs worked out, they'll shoot themselves in the foot because enterprise would never come back. This is why the Xserve is still at a slower processor speed than the desktop (since Apple is overclocking the desktop machines).
All that being said, that's just my opinion, and Apple could very well do the opposite; however, I believe that you have to be very humble and conservative competing in the server world (unlike NT server).
<strong>
I tend to disagree with people that Apple would intro the 970 in the Xserve because of the Xserve market. The Xserve is Apple's attempt to get into server rooms. This is an environment that shares no relation to desktop and reliability/uptime is crucial. I would think that such a shift in processors would be allowed to mature a few months in the desktop market before selling an Xserve with it. Apple is trying to win new enterprise business and if they ship a 1st gen server without the bugs worked out, they'll shoot themselves in the foot because enterprise would never come back. This is why the Xserve is still at a slower processor speed than the desktop (since Apple is overclocking the desktop machines).
All that being said, that's just my opinion, and Apple could very well do the opposite; however, I believe that you have to be very humble and conservative competing in the server world (unlike NT server).</strong><hr></blockquote>
I agree, but for a different reason. If Apple introduces the PPC970 on the Xserve before the PowerMac, what happens to PowerMac sales in the interim? No one can be certain, but I would bet that it would not be pretty.
<strong>What Power Mac sales?</strong><hr></blockquote>
They are still selling ~150k per quarter.
<strong>
Now that's funny. Boy would that burn Microsoft's butt. <img src="graemlins/lol.gif" border="0" alt="[Laughing]" />
Just wondering if it's possible though? Marklar??
[ 02-01-2003: Message edited by: rickag ]</strong><hr></blockquote>
I think in 2001 in December Macosrumors ran a few crazy reports where they claimed the Moto G5 chip would make virtual pc run faster then normal Windows on a Dell.
Basically this will never happen. Maybe with a 5 gzh 970 but at that time Intel will have a faster chip also.
<strong>
Now that's funny. Boy would that burn Microsoft's butt. <img src="graemlins/lol.gif" border="0" alt="[Laughing]" />
Just wondering if it's possible though? Marklar??
[ 02-01-2003: Message edited by: rickag ]</strong><hr></blockquote>
I think in 2001 in December Macosrumors ran a few crazy reports where they claimed the Moto G5 chip would make virtual pc run faster then normal Windows on a Dell.
Basically this will never happen. Maybe with a 5 gzh 970 but at that time Intel will have a faster chip also.