Changing the CPU architecture for a mass-market product isn't like changing Nokia Face Plates.
Even if they did and could get it to run X that would require HUGE amounts of dollars and work.
If they did, it would have to support all legacy stuff ala Itanium does for x86. And we all know that is catching on like wildfire!
Zaz</strong><hr></blockquote>
Apple has the money. If the new architecture is fast enough, it could emulate the old architecture faster than the fastest old machines.
But there are still problems, such as asking developers it rewrite/optimize their code again (after the Great OSX Carbon/Cocoa rewrite), and having different architectures on different current Macs.
But Apple could solve this with a hardware PowerPC translation unit, similar to how the current Athlons work(RISC core+x86 translator).
The G4 aint dead yet, it's just lying there bleeding a little, needs a brain implant and some steroids to get it up to strength. MOTO designs some good stuff. I would think that Apple has been aware of this and probably has been getting MOTO to design a chip for them specifically. That would be a good reason for MoTo to remove the G5 (7500) from their roadmap, since it's built with apple money. This would also fit with MoTo saying they are sticking to the MPX bus for THEIR stuff.
Time will tell and there ain't nobody talking about it that knows what it is. Dorsal's stuff sounds good but who knows if it was research stuff (which all companies do) or working prototypes.
If it ain't NY 2002, then anytime after when the stuff is fully tested, Apple will release it. They would be dumb to curtail their sales just to do something at a stupid expo (they don't make money on expos, they make money selling (hardware and software)
Something will be coming and nobody knows when. Apple may know, but problems come up and I would bet that they ain't gonna let another 500Mhz fiasco happen. That would kill them.
The G5 will be here soon, but I fear that when it finally arrives it will not be much to write home about. If it had been released in January, then it would have been a very competitive chip, but the longer Moto waits the more obsolete it becomes. At the rate things are going, the G5 will be antiquated tech upon arrival. Who's really going to care about a 1.5 GHz G5 when the Pentium 4 is clocked at over 3 GHz? Sure the G5 will probably be more efficient, but twice as efficient? Intel is using brute GHz to stomp the competition, and it's working.
Mhz don't mean anything to me. Bandwidth is what counts. If you can't move it in and out of the processor the processor can't do any thing to it. DDR won't cut it. It's time for a new approach and bandwidth on the board and processors is what's needed. The Starship Enterprise ain't gonna run on a 10Ghz Intel x86 chip. (of course the Enterprise will have the necessary power to do it but it will limit speeds to Warp 5 only)
It' s time to think out of the box and move towards some new technology. A couple of G4+ @ 1.5 GHZ with DDR will work for awhile but it's time to move forward. This is what MoTo did originally with the 8088 and risc and was what they were good at at one time (and still are in the embedded world)
<strong>This is what MoTo did originally with the 8088 and risc and was what they were good at at one time (and still are in the embedded world)</strong><hr></blockquote>
The 8088 is an Intel processor. You're probably referring to Moto's 88000 or 88k line.
I remember reading a speculative piece in Scientific American a few years ago that in the future CPUs will seem more like scattered processing units floating in a sea of RAM. Scaling up from the chip to the mobo, it seems to me that UMA architectures (pardon the tautology) along the lines of what Silicon Graphics have been doing since the O2 is a big step in this direction.
Correct me if I'm wrong, but doesn't the xbox have a similar architecture? And for that matter I vaguely recall reading an IBM spokesthing bragging about the funky caching that they've got in the gamecube.
I strongly suspect that your 8 yr old brother is playing Virtua Fighter on a box that looks very much like the future of mobo design.
<strong>Correct me if I'm wrong, but doesn't the xbox have a similar architecture?
</strong><hr></blockquote>
Nope, it doesn't. The X-Box basically uses a version of the nForce modified to work with a Celeron processor (i.e. this version uses Intel's GTL+ FSB rather than AMD's EV6 FSB). It's a classical northbridge-and-southbridge design, with the exception that those two are interconnected by means of an 800MB/s HyperTransport link rather than PCI or some other proprietary interconnect (MuTIOL, V-Link, HubLink).
<strong>I remember reading a speculative piece in Scientific American a few years ago that in the future CPUs will seem more like scattered processing units floating in a sea of RAM. Scaling up from the chip to the mobo, it seems to me that UMA architectures (pardon the tautology) along the lines of what Silicon Graphics have been doing since the O2 is a big step in this direction.
Correct me if I'm wrong, but doesn't the xbox have a similar architecture? And for that matter I vaguely recall reading an IBM spokesthing bragging about the funky caching that they've got in the gamecube.
I strongly suspect that your 8 yr old brother is playing Virtua Fighter on a box that looks very much like the future of mobo design.</strong><hr></blockquote>
I remember reading that. The author went out of his way to say that he had almost no idea where CPUs would go but since they wanted him to guess....
In short he guess that more on CPU cache would be in the future. Not very wild speculation.
"The G5 will be here soon, but I fear that when it finally arrives it will not be much to write home about. If it had been released in January, then it would have been a very competitive chip, but the longer Moto waits the more obsolete it becomes. At the rate things are going, the G5 will be antiquated tech upon arrival. Who's really going to care about a 1.5 GHz G5 when the Pentium 4 is clocked at over 3 GHz? Sure the G5 will probably be more efficient, but twice as efficient? Intel is using brute GHz to stomp the competition, and it's working. "
I think actually you are thinking of Mot's 68k line (680x0). I used to have a blazing 33Mhz Quadra on a 68040.
`Death to the murderers.`</strong><hr></blockquote>
No, I'm pretty sure he meant the 88000. This was Motorola's RISC design of the day and it was much more powerful than it's 680x0 cousin as I recall. I recall reading speculation of a "Monster Mac" using the 88000 series by Dvorak of all people.
I thought that this year, Apple and Moto may take the fight to x86.
But the whole rate of progress seems very leisurely.
Apple seems to capitulate in the 'megahurts'.
A little ray of hope is the 7500 (G5) with Rapid Io. Broader. Throughput.
Maybe this gives Apple a chance to level the playing fields.
But I still think it's going to have to go some against the AMD 'Hammer'.
The comparitive benches next January should be very interesting. Still, I don't think it (the 7500) will be any slouch. Can't be any worse than a PC World Athlon 2.1 xp kicking the crap out of Apple's low end 'power'Mac...
If it turns out to be the G5, I'm buying. If it's just a ramped G4 in 'Rapid Io' boots...then...I don't know.
I was kinda hopin', like the original G4 announcement that Steve Jobs would surprise us all with a 'G5' annoucement.
I'm with qazII and Scott F on this one. I think what apple needs is to employ some bright young things who are not neccessarily chip deigners and engineers to sit ina room day after day and say things like:
"what if . . ."
Then let the engineers and tech whizzes try to solve the problems put to them. What apple needs is an insanely great change in how we think about processor design. For example, does a computer really need a single CPu responsible for evrything? Can a motherboard be a CPU if you like with its various "engines" functioning as a system?
At the mo it seems as if apple just employs a load of people whose sole job is to say "we can't becuase . . ." There must be people left in the world who no longer believe the earth is flat.
Apple doesn't have the money to implement a new paradigm in CPU/mobo architecture. Get real, they already have to price their computers at twice that of Wintels just to remain profitable. There's a reason for Macs having antiquated tech, it's that Apple doesn't have the money to keep their computers up to date.
No, I'm pretty sure he meant the 88000. This was Motorola's RISC design of the day and it was much more powerful than it's 680x0 cousin as I recall. I recall reading speculation of a "Monster Mac" using the 88000 series by Dvorak of all people.
Steve</strong><hr></blockquote>
<off-topic> True, that was Motorola's RISC line. It was said to be quite fast by some. There wasn't too much speculation about the line going into Macs, from what I can remember. At the same time the 88000 series as coming to fruition, Apple decided to go with the PPC. As consolation to Motorola, they preserved the 88000's bus design by putting it into the PPC 601.
At the same time, Motorola was also completing the 68060, the successor to the 68040. It was roughly equivalent in performance to the 601 but was larger and ran hotter. </off-topic>
Apple must realize it cannot continue to allow the MHz gap to widen, MHz Myth or otherwise. They surely have an answer, whether it means taking development in-house, going with IBM completely or going with a different player. If Apple had to switch chip lines in order to remain in business, it would. Any switch of that magnitude would have to include PPC instruction set compatibility, since another code transition at this point in time would be a disaster.
Comments
<strong>
3 words:
Legacy Code Base
Changing the CPU architecture for a mass-market product isn't like changing Nokia Face Plates.
Even if they did and could get it to run X that would require HUGE amounts of dollars and work.
If they did, it would have to support all legacy stuff ala Itanium does for x86. And we all know that is catching on like wildfire!
Zaz</strong><hr></blockquote>
Apple has the money. If the new architecture is fast enough, it could emulate the old architecture faster than the fastest old machines.
But there are still problems, such as asking developers it rewrite/optimize their code again (after the Great OSX Carbon/Cocoa rewrite), and having different architectures on different current Macs.
But Apple could solve this with a hardware PowerPC translation unit, similar to how the current Athlons work(RISC core+x86 translator).
Time will tell and there ain't nobody talking about it that knows what it is. Dorsal's stuff sounds good but who knows if it was research stuff (which all companies do) or working prototypes.
If it ain't NY 2002, then anytime after when the stuff is fully tested, Apple will release it. They would be dumb to curtail their sales just to do something at a stupid expo (they don't make money on expos, they make money selling (hardware and software)
Something will be coming and nobody knows when. Apple may know, but problems come up and I would bet that they ain't gonna let another 500Mhz fiasco happen. That would kill them.
It' s time to think out of the box and move towards some new technology. A couple of G4+ @ 1.5 GHZ with DDR will work for awhile but it's time to move forward. This is what MoTo did originally with the 8088 and risc and was what they were good at at one time (and still are in the embedded world)
<strong>
The MPX bus is not new, but is Mot's proprietary CPU-to-memory controller bus that is part of the G3/4 chips.
</strong><hr></blockquote>
Actually, it's G4 only. The G3 only supports the older 60x bus.
Bye,
RazzFazz
<strong>This is what MoTo did originally with the 8088 and risc and was what they were good at at one time (and still are in the embedded world)</strong><hr></blockquote>
The 8088 is an Intel processor. You're probably referring to Moto's 88000 or 88k line.
Bye,
RazzFazz
Correct me if I'm wrong, but doesn't the xbox have a similar architecture? And for that matter I vaguely recall reading an IBM spokesthing bragging about the funky caching that they've got in the gamecube.
I strongly suspect that your 8 yr old brother is playing Virtua Fighter on a box that looks very much like the future of mobo design.
<strong>Correct me if I'm wrong, but doesn't the xbox have a similar architecture?
</strong><hr></blockquote>
Nope, it doesn't. The X-Box basically uses a version of the nForce modified to work with a Celeron processor (i.e. this version uses Intel's GTL+ FSB rather than AMD's EV6 FSB). It's a classical northbridge-and-southbridge design, with the exception that those two are interconnected by means of an 800MB/s HyperTransport link rather than PCI or some other proprietary interconnect (MuTIOL, V-Link, HubLink).
Bye,
RazzFazz
[ 06-03-2002: Message edited by: RazzFazz ]</p>
<strong>I remember reading a speculative piece in Scientific American a few years ago that in the future CPUs will seem more like scattered processing units floating in a sea of RAM. Scaling up from the chip to the mobo, it seems to me that UMA architectures (pardon the tautology) along the lines of what Silicon Graphics have been doing since the O2 is a big step in this direction.
Correct me if I'm wrong, but doesn't the xbox have a similar architecture? And for that matter I vaguely recall reading an IBM spokesthing bragging about the funky caching that they've got in the gamecube.
I strongly suspect that your 8 yr old brother is playing Virtua Fighter on a box that looks very much like the future of mobo design.</strong><hr></blockquote>
I remember reading that. The author went out of his way to say that he had almost no idea where CPUs would go but since they wanted him to guess....
In short he guess that more on CPU cache would be in the future. Not very wild speculation.
I think actually you are thinking of Mot's 68k line (680x0). I used to have a blazing 33Mhz Quadra on a 68040.
`Death to the murderers.`
[ 06-03-2002: Message edited by: Dead Member ]</p>
Something like that.
Lemon Bon Bon
<strong>
I think actually you are thinking of Mot's 68k line (680x0). I used to have a blazing 33Mhz Quadra on a 68040.
`Death to the murderers.`</strong><hr></blockquote>
No, I'm pretty sure he meant the 88000. This was Motorola's RISC design of the day and it was much more powerful than it's 680x0 cousin as I recall. I recall reading speculation of a "Monster Mac" using the 88000 series by Dvorak of all people.
Steve
But the whole rate of progress seems very leisurely.
Apple seems to capitulate in the 'megahurts'.
A little ray of hope is the 7500 (G5) with Rapid Io. Broader. Throughput.
Maybe this gives Apple a chance to level the playing fields.
But I still think it's going to have to go some against the AMD 'Hammer'.
The comparitive benches next January should be very interesting. Still, I don't think it (the 7500) will be any slouch. Can't be any worse than a PC World Athlon 2.1 xp kicking the crap out of Apple's low end 'power'Mac...
If it turns out to be the G5, I'm buying. If it's just a ramped G4 in 'Rapid Io' boots...then...I don't know.
I was kinda hopin', like the original G4 announcement that Steve Jobs would surprise us all with a 'G5' annoucement.
Lemon Bon Bon <img src="graemlins/hmmm.gif" border="0" alt="[Hmmm]" />
Dual 1.33
Dual 1.2?
1 gig?
Something like that? Give or take the odd multiplyer quibble?
Well, at least Quartz Extreme looks exciting. Maybe a case re
I've got the horrible feeling they'll be 'leisurely' updates.
Lemon Bon Bon
"what if . . ."
Then let the engineers and tech whizzes try to solve the problems put to them. What apple needs is an insanely great change in how we think about processor design. For example, does a computer really need a single CPu responsible for evrything? Can a motherboard be a CPU if you like with its various "engines" functioning as a system?
At the mo it seems as if apple just employs a load of people whose sole job is to say "we can't becuase . . ." There must be people left in the world who no longer believe the earth is flat.
<strong>
No, I'm pretty sure he meant the 88000. This was Motorola's RISC design of the day and it was much more powerful than it's 680x0 cousin as I recall. I recall reading speculation of a "Monster Mac" using the 88000 series by Dvorak of all people.
Steve</strong><hr></blockquote>
<off-topic> True, that was Motorola's RISC line. It was said to be quite fast by some. There wasn't too much speculation about the line going into Macs, from what I can remember. At the same time the 88000 series as coming to fruition, Apple decided to go with the PPC. As consolation to Motorola, they preserved the 88000's bus design by putting it into the PPC 601.
At the same time, Motorola was also completing the 68060, the successor to the 68040. It was roughly equivalent in performance to the 601 but was larger and ran hotter. </off-topic>
Apple must realize it cannot continue to allow the MHz gap to widen, MHz Myth or otherwise. They surely have an answer, whether it means taking development in-house, going with IBM completely or going with a different player. If Apple had to switch chip lines in order to remain in business, it would. Any switch of that magnitude would have to include PPC instruction set compatibility, since another code transition at this point in time would be a disaster.
[ 06-04-2002: Message edited by: Big Mac ]</p>