depends on how fast they can advance each process...if they quickly get to 2.5 gHz on the .13 process and then advance that a step or two...maybe to 2.8 or even 3.0, then they move to a .09 process in early 2004 and that gives an immediate boost of 30%....suddenly it may not seem impossible to be at 4 or 4.5 by the end of 2004....seems insane to me, but isn't FH for the insane anyway?
If the 970/980 would reach 4.5ghz by the end of 2004, that would mean more than doubling the performance in about a year. I doubt very much that we will see this kind of speed increase in the first year. Heck, I mean just going from 1.8ghz this july/august to about 3 ghz in a year would be very impressive. But 4.5 ghz ???
Pentium4 did it -- started at 1.4 GHz and is now at 3 GHz, heading up. IBM started considerably later, but as a result gains the advantage of having the more advanced processes more quickly. The pipelines on the 970 are roughly as long as those on the Pentium4, and we don't yet know enough about other internal details to have any idea how it will scale compared to the Intel design. The 4.5 GHz number is for the 9x0 chips, after the POWER5 arrives which is probably more than a year from now. The fact that the POWER4 has gained 70% in 2.5 years shows that IBM is capable of scaling quickly, even on a design intended more for reliability and robustness.
If IBM can get a 2.5 gig 970 by the end 03 beg' 04...? That will be a 5 gig Pentium 4. That's twice per clock. That's going to run Intel really hard. There's no WAY they're going to have a five gig P4 out the door by the end of the year/early 2004? And if the 980 is anything like the 'four times the performance' muted in the Power 5 over Power 4...then...at 4.5 gig, a 980 is going to get all medieval over the Prescott hurtz!!! Equivalent of more than 9 gig of P4 performance! That's in single formations! And, the 980 ought to do far more per clock! So that's conservative.
That's really aggressive ramping. The anticipation is mouth-watering.
The next year and a half/ish...things are going to be very interesting. I can't wait for the can of whoop-ass IBM's going to open all over Intel.
If IBM can get a 2.5 gig 970 by the end 03 beg' 04...? That will be a 5 gig Pentium 4. That's twice per clock. That's going to run Intel really hard. There's no WAY they're going to have a five gig P4 out the door by the end of the year/early 2004? And if the 980 is anything like the 'four times the performance' muted in the Power 5 over Power 4...then...at 4.5 gig, a 980 is going to get all medieval over the Prescott hurtz!!! Equivalent of more than 9 gig of P4 performance!
THREE ... TWO ... ONE ... ZERO
And we have a lift off!
Calm down. We all like that shiny future these wild rumors give our favorit toy but you shouldn't loose the contact to the ground. Technically nothing is known about the POWER5 and its little brother 980 (one core) or not so little brother 9800 (two cores).
You might not buy it, but lots of the Pro market would. High-end PowerMac / Xstation, it doesn't really matter -- the same principle applies. Apple has had an "Ultimate" configuration on their store for a long time, but it has never really had "an edge".
Pros can by as much rumors as they want
Why should Apple do a Xstation, when IBM delivers so powerfull cpus over the next years that the 'normal' powermac will eat the workstation market alive?
I don't like the "Ultimate" configuration too but as it is today it doesn't cost Apple a penny investment. so from the view of Apple it's perfect
Considering the POWER4 and POWER4+ have scaled well (1.1GHz->1.7GHz) in a short amount of time, just over a year, the 970, a smaller version of the POWER4 with a single core and a third of the L2 cache, should have a sharper incline in MHz. In fact 2.0GHz max speed right on launch should not be surprising. Lets suppose in August the 970 Macs are released with a 2GHz top speed. About that time IBM should be having higher numbers of 2.2-2.5GHz 970s. Apple stockpiles them until late 2003 when in December or January they release the 2.0-2.5GHz 970 machines. Still based on 130nm. At this point IBM is focusing all their attention on the 90nm 970 parts that should start at 2.5GHz and scale at first up to 2.8GHz or maybe a few 3GHz parts. They would be ready in time for a may-June release leaving 90nm 970s in late 2004 scaling up to 3.5GHz. Not too shabby but nowhere close to the 4.5GHz quoted speed. Remember the drop from 180nm to 130nm is steeper that 130nm to 90nm and the Power4 only increased in MHz about 400MHz (when the 1.3GHz POWER4 was max and the current POWER4+ is 1.7GHz). Thats about a 30% increase (to be expected) so a similar increase with the 970 max (assuming they can eek out 2.5GHz from the 130nm 970) would be 3.25GHz. But this is not counting any other process and technical improvements that IBM may introduce with the 90nm part. So 3.5-4GHz may be a reality for late 2004 for the 970.
If the MacBidouille correspondent is right, and we're looking at 130nm -> 130nm + SiLK -> 90nm ->90nm + (something like SiLK), then I can easily see the 970 ramping smoothly up to very high frequencies. IBM has a neat little roadmap based on proven tech to take the 970 into 2004.
As Programmer has pointed out, it's pipelined about as deeply as the Pentium 4. It should have no trouble scaling up.
This is not the G4, people.
As for the ultra-high-end workstations, I'm going to have to continue being a wet blanket here. I just don't see Apple getting into the $10K+ market. I can see one or two of two things happening: 1) Apple releases really hot machines within the PowerMac price range, and counts on high-speed interconnects and efficient clustering/distributed computing software to scale up, and/or 2) Apple licenses IBM to build boxes for markets that Apple itself is not interested in focusing on, running Mac OS X Server. IBM knows big iron, and customers know that IBM knows big iron.
I think that Core Computers points to the viability of 2), oddly enough. I'm surprised that they're still up, and I'm more surprised that Larry of OWC has publicly backed them, given that OWC really wants to stay in Apple's good graces. I think this points to the fact that Apple is willing to entertain the idea of non-Apple Mac-compatible suppliers on an individual basis, if they're tackling markets that Apple itself is not interested in. This avoids all the pitfalls of the old clone program. Core can handle the tweaker/Matsu market, IBM can handle their traditional market, and both can run Mac OS X. Apple, meanwhile, can do what they've always done without fear of someone aiming for their bread-and-butter hardware. This will accomplish what the clone program did not, by expanding the appeal of the platform into markets that Apple has not (and will not) enter themselves. If they do it carefully, it can work. And the 970 can realize its full potential.
I can image the CAD/CAE boys would be interested to see a high-performance 64 bit, 4-way system with hardware accelerated X11. I wonder if Apple's talked to MSC or Dassault...
Well, this CAE boy is definitely interested in MSC and Dassault getting on the Mac OSX platform. Particularly with the Altivec unit's potential in this regard. A vector processor is exactly the thing MSC could make a lot of use of. (And isn't nastran predominantly single precision?)
I can't remember the last time Apple was in such a truly revolutionary position. The possibilities are incredible.
Come to think if it, I don't think they have ever been in this good a position. They have done some amazing things over the years, but I can't think of a single instance where hardware, software, OS and all the other miscellaneous pieces have come together at the same time.
This summer and going into next year, is going to be a lot of fun.
Well, this CAE boy is definitely interested in MSC and Dassault getting on the Mac OSX platform. Particularly with the Altivec unit's potential in this regard. A vector processor is exactly the thing MSC could make a lot of use of. (And isn't nastran predominantly single precision?)
MM
It does both single- and double-precision, depending. A lot could be made of the single-precision Altivec support considering that a lot of FEA code can be efficiently parallellized, though I can't guess as to how much time it spends working on single-precision (yet another point for double-precision Altivec).
NASTRAN could certainly use the larger address space. It certainly could benefit from the solid FPU. It certainly can benefit from dual and quad systems. It can benefit from Altivec.
The real question is would it make sense to bring it down from RS/6000 or over from x86 to the 970? Bringing it down from IBM workstations gets you much cheaper workstations without much sacrifice provided that Apple doesn't waste that FSB and builds boxes worthy of the task and might even benefit things with Altivec. Coming across from x86, Altivec helps, but the combination of OS X/970 has to provide something that Windows/Opteron can't, and that *might* be long-term security, oddly enough.
I can't imagine that any of the CAD companies are ignoring a mainstream (read: cheap) 64bit unix solution. The biggest obstacle in all of this is that it's *Apple* and not, well, almost anybody else. There's something unsettling about jumpin on a platform as a premiere low-cost CAD workstation that just a few months back rocked the world by launching an online music store.
I can't imagine that any of the CAD companies are ignoring a mainstream (read: cheap) 64bit unix solution. The biggest obstacle in all of this is that it's *Apple* and not, well, almost anybody else. There's something unsettling about jumpin on a platform as a premiere low-cost CAD workstation that just a few months back rocked the world by launching an online music store.
And people wonder why I suggest that Apple might license MacOS X to IBM.
Wasn't the Power5 (and I assume the 980) suppose to have some form of hyperthreading? Can't seem to locate any info and don't think that was the term they used.
As for the ultra-high-end workstations, I'm going to have to continue being a wet blanket here. I just don't see Apple getting into the $10K+ market. I can see one or two of two things happening: 1) Apple releases really hot machines within the PowerMac price range, and counts on high-speed interconnects and efficient clustering/distributed computing software to scale up, and/or 2) Apple licenses IBM to build boxes for markets that Apple itself is not interested in focusing on, running Mac OS X Server. IBM knows big iron, and customers know that IBM knows big iron.
Well, I'm glad somebody who actually knows a thing or two is playing the Wet Blanket? around here; but still, I can certainly see Apple doing something more "Workstationesque" for the FCP4 set - especially since the architecture of FCP4 is designed to be as real time as possible. After all, the market this software is aimed at has no qualms about dropping $10K on a piece of gear - especially if it helps to speed up the most time consuming process of all - editing.
I don't know about $60,000 64way machines, after all, Shake is probably designed to run on clusters, not massively parallel single machines (sexy though it would be to have real time compositing) - but I sure can see 4 or maybe even 8 way workstations if they can come in at around the $10K to 15K price point.
And people wonder why I suggest that Apple might license MacOS X to IBM.
Licencing MacOS X to only IBM would be a masterstroke. It gives the enterprise market a venerable alternative to purchasing Apple hardware as a platform on which to run OS X - and one that any IT head could sign-off and sleep comfortably at night.
And people wonder why I suggest that Apple might license MacOS X to IBM.
Oh, I don't wonder one bit. I've been thinking about that since OS X first showed up. Several things make me think that it may yet come:
1) Apple hasn't had the whole set of goods in a while. OS X on less than inappropriate hardware is tough to sell. The Xserve opened a big door there. The 970 will open more.
2) I doubt that IBM would have gone for something like the 970 back in the OS 9 days. They must be getting something more out of this chip than just a regular check from Apple (though we shouldn't underestimate IBMs happiness with nothing more than a regular check). The 970 to me represents acknowledgement from IBM that either Apple has something valuable to IBM - such as an OS X license, or that IBM sees a market opportunity here to sell low-end 64 bit unix workstations that MS/AMD/Intel may miss but which Apple can ride along with.
Lets face it, though. We've had 64 bit workstations with how many unix OSes and variants now? And how many are still out there? How many has IBM supported? OS X is something different and special here and they might see that.
3) Steve doesn't object to licensing the OS. He objects to licensing the OS to people that will compete with Apple. IBM won't do that. Even if they do sell the same hardware, it'll be a whole other market with a whole other focus. As much as IBM doesn't appeal to consumers, Apple doesn't appeal to enterprise.
4) Apple just dropped the price points significantly. For those of you expecting a $5000 970 - you're wrong. Apple pretty much keeps price points steady and simply improves the package for the price - that's what all of those nutty deals are for RAM, Office, etc. even for brand-new systems. Apple isn't blowing out excess hardware - they're sweetening the pot so they don't have to move price points. Don't forget what happens when Apple raises prices - even a little. The large drop in price points reflects Apple ramping up to compete more broadly, and the shift of profits from the core desktop/laptop line to software and fringe products (iPod, music, etc.) January/February product revs and price drops was a warning shot for everybody that big things were coming in summer:
"Now Apple makes almost as much operating profit on each iPod it sells as it does on each iMac, even though the iPod costs a fraction as much to manufacture." Fortune article
The more Apple isn't dependent on the desktop, laptop, and server lines for profits, the more licensing makes sense. More and more Apple can compete with Dell by operating like Microsoft. So, once Apple cuts it's profit margins on conventional hardware to the 10% range, licensing can turn on with minimal risk. Even if IBM does sell a consumer box, Apple will still rake it in with iPods, .mac, iLife, and everything else.
OS X 10.3 + X11 + 970 = something IBM can finally use. I think the license was part of the deal for getting the 970 in our hands. At least I dearly hope so...
Well, I'm glad somebody who actually knows a thing or two is playing the Wet Blanket? around here; but still, I can certainly see Apple doing something more "Workstationesque" for the FCP4 set - especially since the architecture of FCP4 is designed to be as real time as possible. After all, the market this software is aimed at has no qualms about dropping $10K on a piece of gear - especially if it helps to speed up the most time consuming process of all - editing.
I don't know about $60,000 64way machines, after all, Shake is probably designed to run on clusters, not massively parallel single machines (sexy though it would be to have real time compositing) - but I sure can see 4 or maybe even 8 way workstations if they can come in at around the $10K to 15K price point.
Comments
g
Originally posted by NETROMac
If the 970/980 would reach 4.5ghz by the end of 2004, that would mean more than doubling the performance in about a year. I doubt very much that we will see this kind of speed increase in the first year. Heck, I mean just going from 1.8ghz this july/august to about 3 ghz in a year would be very impressive. But 4.5 ghz ???
Pentium4 did it -- started at 1.4 GHz and is now at 3 GHz, heading up. IBM started considerably later, but as a result gains the advantage of having the more advanced processes more quickly. The pipelines on the 970 are roughly as long as those on the Pentium4, and we don't yet know enough about other internal details to have any idea how it will scale compared to the Intel design. The 4.5 GHz number is for the 9x0 chips, after the POWER5 arrives which is probably more than a year from now. The fact that the POWER4 has gained 70% in 2.5 years shows that IBM is capable of scaling quickly, even on a design intended more for reliability and robustness.
If IBM can get a 2.5 gig 970 by the end 03 beg' 04...? That will be a 5 gig Pentium 4. That's twice per clock. That's going to run Intel really hard. There's no WAY they're going to have a five gig P4 out the door by the end of the year/early 2004? And if the 980 is anything like the 'four times the performance' muted in the Power 5 over Power 4...then...at 4.5 gig, a 980 is going to get all medieval over the Prescott hurtz!!! Equivalent of more than 9 gig of P4 performance! That's in single formations! And, the 980 ought to do far more per clock! So that's conservative.
That's really aggressive ramping. The anticipation is mouth-watering.
The next year and a half/ish...things are going to be very interesting. I can't wait for the can of whoop-ass IBM's going to open all over Intel.
PPC. Has your time come..?
Lemon Bon Bon
Originally posted by Lemon Bon Bon
If IBM can get a 2.5 gig 970 by the end 03 beg' 04...? That will be a 5 gig Pentium 4. That's twice per clock. That's going to run Intel really hard. There's no WAY they're going to have a five gig P4 out the door by the end of the year/early 2004? And if the 980 is anything like the 'four times the performance' muted in the Power 5 over Power 4...then...at 4.5 gig, a 980 is going to get all medieval over the Prescott hurtz!!! Equivalent of more than 9 gig of P4 performance!
THREE ... TWO ... ONE ... ZERO
And we have a lift off!
Calm down. We all like that shiny future these wild rumors give our favorit toy but you shouldn't loose the contact to the ground. Technically nothing is known about the POWER5 and its little brother 980 (one core) or not so little brother 9800 (two cores).
Originally posted by Programmer
You might not buy it, but lots of the Pro market would.
Pros can by as much rumors as they want
Why should Apple do a Xstation, when IBM delivers so powerfull cpus over the next years that the 'normal' powermac will eat the workstation market alive?
I don't like the "Ultimate" configuration too but as it is today it doesn't cost Apple a penny investment. so from the view of Apple it's perfect
As Programmer has pointed out, it's pipelined about as deeply as the Pentium 4. It should have no trouble scaling up.
This is not the G4, people.
As for the ultra-high-end workstations, I'm going to have to continue being a wet blanket here. I just don't see Apple getting into the $10K+ market. I can see one or two of two things happening: 1) Apple releases really hot machines within the PowerMac price range, and counts on high-speed interconnects and efficient clustering/distributed computing software to scale up, and/or 2) Apple licenses IBM to build boxes for markets that Apple itself is not interested in focusing on, running Mac OS X Server. IBM knows big iron, and customers know that IBM knows big iron.
I think that Core Computers points to the viability of 2), oddly enough. I'm surprised that they're still up, and I'm more surprised that Larry of OWC has publicly backed them, given that OWC really wants to stay in Apple's good graces. I think this points to the fact that Apple is willing to entertain the idea of non-Apple Mac-compatible suppliers on an individual basis, if they're tackling markets that Apple itself is not interested in. This avoids all the pitfalls of the old clone program. Core can handle the tweaker/Matsu market, IBM can handle their traditional market, and both can run Mac OS X. Apple, meanwhile, can do what they've always done without fear of someone aiming for their bread-and-butter hardware. This will accomplish what the clone program did not, by expanding the appeal of the platform into markets that Apple has not (and will not) enter themselves. If they do it carefully, it can work. And the 970 can realize its full potential.
Originally posted by johnsonwax
I can image the CAD/CAE boys would be interested to see a high-performance 64 bit, 4-way system with hardware accelerated X11. I wonder if Apple's talked to MSC or Dassault...
Well, this CAE boy is definitely interested in MSC and Dassault getting on the Mac OSX platform.
MM
Originally posted by Outsider
Considering the POWER4 and POWER4+ have scaled well (1.1GHz->1.7GHz) in a short amount of time, just over a year
Woops, you're right -- I counted an extra year in there somehow.
So 3.5-4GHz may be a reality for late 2004 for the 970.
I was thinking early 2005 for a 980 at ~4 - 4.5 GHz, so we're not really that far apart.
I can't remember the last time Apple was in such a truly revolutionary position. The possibilities are incredible.
Come to think if it, I don't think they have ever been in this good a position. They have done some amazing things over the years, but I can't think of a single instance where hardware, software, OS and all the other miscellaneous pieces have come together at the same time.
This summer and going into next year, is going to be a lot of fun.
Originally posted by MartianMatt
Well, this CAE boy is definitely interested in MSC and Dassault getting on the Mac OSX platform.
MM
It does both single- and double-precision, depending. A lot could be made of the single-precision Altivec support considering that a lot of FEA code can be efficiently parallellized, though I can't guess as to how much time it spends working on single-precision (yet another point for double-precision Altivec).
NASTRAN could certainly use the larger address space. It certainly could benefit from the solid FPU. It certainly can benefit from dual and quad systems. It can benefit from Altivec.
The real question is would it make sense to bring it down from RS/6000 or over from x86 to the 970? Bringing it down from IBM workstations gets you much cheaper workstations without much sacrifice provided that Apple doesn't waste that FSB and builds boxes worthy of the task and might even benefit things with Altivec. Coming across from x86, Altivec helps, but the combination of OS X/970 has to provide something that Windows/Opteron can't, and that *might* be long-term security, oddly enough.
I can't imagine that any of the CAD companies are ignoring a mainstream (read: cheap) 64bit unix solution. The biggest obstacle in all of this is that it's *Apple* and not, well, almost anybody else. There's something unsettling about jumpin on a platform as a premiere low-cost CAD workstation that just a few months back rocked the world by launching an online music store.
Originally posted by johnsonwax
I can't imagine that any of the CAD companies are ignoring a mainstream (read: cheap) 64bit unix solution. The biggest obstacle in all of this is that it's *Apple* and not, well, almost anybody else. There's something unsettling about jumpin on a platform as a premiere low-cost CAD workstation that just a few months back rocked the world by launching an online music store.
And people wonder why I suggest that Apple might license MacOS X to IBM.
Originally posted by Programmer
I was thinking early 2005 for a 980 at ~4 - 4.5 GHz, so we're not really that far apart. [/B]
Yeah and then watch IBM pull a Moto on us and not upgrade the 4500 Mhz chip to 5000 Mhz for another year.
Originally posted by Amorph
As for the ultra-high-end workstations, I'm going to have to continue being a wet blanket here. I just don't see Apple getting into the $10K+ market. I can see one or two of two things happening: 1) Apple releases really hot machines within the PowerMac price range, and counts on high-speed interconnects and efficient clustering/distributed computing software to scale up, and/or 2) Apple licenses IBM to build boxes for markets that Apple itself is not interested in focusing on, running Mac OS X Server. IBM knows big iron, and customers know that IBM knows big iron.
Well, I'm glad somebody who actually knows a thing or two is playing the Wet Blanket? around here; but still, I can certainly see Apple doing something more "Workstationesque" for the FCP4 set - especially since the architecture of FCP4 is designed to be as real time as possible. After all, the market this software is aimed at has no qualms about dropping $10K on a piece of gear - especially if it helps to speed up the most time consuming process of all - editing.
I don't know about $60,000 64way machines, after all, Shake is probably designed to run on clusters, not massively parallel single machines (sexy though it would be to have real time compositing) - but I sure can see 4 or maybe even 8 way workstations if they can come in at around the $10K to 15K price point.
Hell, I used to sell $7K+ MacIIfx's
Originally posted by Programmer
And people wonder why I suggest that Apple might license MacOS X to IBM.
Licencing MacOS X to only IBM would be a masterstroke. It gives the enterprise market a venerable alternative to purchasing Apple hardware as a platform on which to run OS X - and one that any IT head could sign-off and sleep comfortably at night.
Originally posted by Programmer
And people wonder why I suggest that Apple might license MacOS X to IBM.
Oh, I don't wonder one bit. I've been thinking about that since OS X first showed up. Several things make me think that it may yet come:
1) Apple hasn't had the whole set of goods in a while. OS X on less than inappropriate hardware is tough to sell. The Xserve opened a big door there. The 970 will open more.
2) I doubt that IBM would have gone for something like the 970 back in the OS 9 days. They must be getting something more out of this chip than just a regular check from Apple (though we shouldn't underestimate IBMs happiness with nothing more than a regular check). The 970 to me represents acknowledgement from IBM that either Apple has something valuable to IBM - such as an OS X license, or that IBM sees a market opportunity here to sell low-end 64 bit unix workstations that MS/AMD/Intel may miss but which Apple can ride along with.
Lets face it, though. We've had 64 bit workstations with how many unix OSes and variants now? And how many are still out there? How many has IBM supported? OS X is something different and special here and they might see that.
3) Steve doesn't object to licensing the OS. He objects to licensing the OS to people that will compete with Apple. IBM won't do that. Even if they do sell the same hardware, it'll be a whole other market with a whole other focus. As much as IBM doesn't appeal to consumers, Apple doesn't appeal to enterprise.
4) Apple just dropped the price points significantly. For those of you expecting a $5000 970 - you're wrong. Apple pretty much keeps price points steady and simply improves the package for the price - that's what all of those nutty deals are for RAM, Office, etc. even for brand-new systems. Apple isn't blowing out excess hardware - they're sweetening the pot so they don't have to move price points. Don't forget what happens when Apple raises prices - even a little. The large drop in price points reflects Apple ramping up to compete more broadly, and the shift of profits from the core desktop/laptop line to software and fringe products (iPod, music, etc.) January/February product revs and price drops was a warning shot for everybody that big things were coming in summer:
"Now Apple makes almost as much operating profit on each iPod it sells as it does on each iMac, even though the iPod costs a fraction as much to manufacture." Fortune article
The more Apple isn't dependent on the desktop, laptop, and server lines for profits, the more licensing makes sense. More and more Apple can compete with Dell by operating like Microsoft. So, once Apple cuts it's profit margins on conventional hardware to the 10% range, licensing can turn on with minimal risk. Even if IBM does sell a consumer box, Apple will still rake it in with iPods, .mac, iLife, and everything else.
OS X 10.3 + X11 + 970 = something IBM can finally use. I think the license was part of the deal for getting the 970 in our hands. At least I dearly hope so...
Originally posted by jwdawso
JYD - glad to see you apparently have a real life now based on your infrequent contributions.
That was either a compliment or a dig--still trying to figure out which...
Originally posted by OverToasty
Well, I'm glad somebody who actually knows a thing or two is playing the Wet Blanket? around here; but still, I can certainly see Apple doing something more "Workstationesque" for the FCP4 set - especially since the architecture of FCP4 is designed to be as real time as possible. After all, the market this software is aimed at has no qualms about dropping $10K on a piece of gear - especially if it helps to speed up the most time consuming process of all - editing.
I don't know about $60,000 64way machines, after all, Shake is probably designed to run on clusters, not massively parallel single machines (sexy though it would be to have real time compositing) - but I sure can see 4 or maybe even 8 way workstations if they can come in at around the $10K to 15K price point.
Hell, I used to sell $7K+ MacIIfx's
I bought one.