Since the Freescale solution is likely to be very low power and 32-bit, I doubt it will ever be called a "G5". I think the 7448 will remain a "G4". If Freescale comes out with an eye-popping 8641, I can see Apple moving to a different nomenclature to distinguish the mobile/low-end line from the G5/desktop line. Perhaps, M5.
thank you all for bringing up your thought and ideas as well as solid info to the G6-topic. sounds not too bad - i like the stories about dual core and altivec2... at the moment i wouldn't have a problem with a 65nm-process in 2007 as long as other improvements are made....
why in 2007? it's way too long. i want this thing with 2.5 GHz now at 90nm - and i don't think that it should be a real problem for IBM to manufacture something like that... ?
what about graphics? X800 ?
Why are people in these forums so anxious to get a G6? The G5 is a young chip...one reason I want an apple is because they don't become obsolete as fast, but the way people talk on these forums its like as soon as one thing comes out, they want an upgrade. I wouldn't hope to see a G6 until 2007, because I'd like to see them milk the G5 for all its worth, before getting a new chip and building a new OS around it. I plan to get a G5 iMac or eMac(depending on whether or not they upgrade it) at the end of august, once I sell my current PC, and I'd hate for a G6 to be announced right after a buy it. There's something to be said of longevity, and thats why Apple appeals to me.
don't worry - if you decide to buy a g5 iMac now and a week after it there would be a G5 with dual core and altivec2 with an memory-controller-on-chip that would NOT mean that:
- your g5 imac would be outdated
- apple will build a new os around that chip (os x is here to stay for a very long time)
- it would be a bad thing for the pro-market
- i wouldn't be happy with such a machine
- my G4/400 won't run the current OS good enough to be a cool computer
It seems to me that that's the nature of technology. As a consumer, and user, I'd love to own the nicest computer in the world for two years. However, if Apple were to slow down to a lesser pace then that of technological advancement, that would make it more like art, or a club, or something, rather then a computer company. Of course it's purpose is to make money, and while selling technology, they use the best strategy to have a market for the products, while at the same time remaining cutting edge technologically. I'd like to see all the best technology in the world become obsolete every year. I'm always interested in new possibilities. Of course that's extreme. Hypothetically, the computer market would have to change entirely to exist like that, but think of the real benefits and possibilities in our lifetime.
Quote:
Originally posted by Alcibiades
Why are people in these forums so anxious to get a G6? The G5 is a young chip...one reason I want an apple is because they don't become obsolete as fast, but the way people talk on these forums its like as soon as one thing comes out, they want an upgrade. I wouldn't hope to see a G6 until 2007, because I'd like to see them milk the G5 for all its worth, before getting a new chip and building a new OS around it. I plan to get a G5 iMac or eMac(depending on whether or not they upgrade it) at the end of august, once I sell my current PC, and I'd hate for a G6 to be announced right after a buy it. There's something to be said of longevity, and thats why Apple appeals to me.
For me, the interest lies in what all those engineers have up their sleeves. The whole clockspeed race was essentially uninteresting, because nothing changed fundamentally from a systems point of view; it was the same old archaic PC (which I'm using in this post in the general sense of "personal computer") design, just faster, with the occasional interesting bit tagging along for the ride.
Now, things are really getting interesting. Technologies and designs that had been confined to workstations are migrating down to PCs (and farther down, to consoles and other dedicated devices) in earnest. This began a few years ago with the adoption of HyperTransport, and it's accelerating, courtesy of AMD, NVIDIA and IBM especially.
Previously, there was a great deal of pressure on the CPU in PCs simply because high-speed connections elsewhere on the board were expensive, so the CPU had to do everything (the Amiga stands out as an early and noteworthy exception). Now, high-speed connections between chips on the motherboard make it more feasible to farm work out. The massive inequality in clockspeeds between computational units and memory (which used to go the other way, hard as that may be to believe) has made it possible to do some "precalculation" work on data as it streams by, before it gets to the CPU, or the GPU, or wherever the intended destination is (IBM attempted this early in PCs with their Microchannel architecture—but the less said about that, the better).
Dual cores are now feasible; so are multithreaded architectures. So are clusters of independent, loosely coupled processors on a single die (Cell, essentially). The combination of mainstream multitasking and multithreading operating systems (even in handhelds!), an inflection point in the number of transistors available at the right cost, and some other factors that I'm probably missing (well, the constant demands from all sides that will assault a CPU from the various components of a motherboard with high-bandwidth interconnects!), coupled with the most recent obstacles encountered at 90nm and below, lead me to think that we'll be seeing some really ingenious CPU designs in the next few years, popping up in surprising places.
The interesting thing is that this isn't completly out of the question as a future Freescale product if you look at what the other embedded CPU manufacturers are doing. Not saying that it will/must happen, but it's not out of the question.
Anything in particular we should pay close attention to? It is a rather broad industry, after all. I'm not asking for a hint, just a general direction.
Anything in particular we should pay close attention to? It is a rather broad industry, after all.
Maybe he makes reference to the ISSCC, taking place next week in San Francisco. Here is the program. i don't see however why this would be relevant to the Macintosh platform.
i don't see however why this would be relevant to the Macintosh platform.
It might not be; but Programmer was answering my wish to see ingenious new architectures. I don't really care where they end up, I just want to see them.
It's more or less a given that if Apple has anything to show in the next few months (say, WWDC), they've asked the parties responsible for the work to please keep a lid on it.
The only reference, and this could remotely only be relevant, is in page 36:
Quote:
A CELL Processor is a multi-core chip consisting of a 64b Power architecture processor, multiple streaming processors, a flexible IO interface, and a memory interface controller.This SoC is implemented in 90nm SOI technology. The chip is designed with a high degree of modularity and reuse to maximize the custom circuit content and achieve a high-frequency clock-rate.
Another topic, of interest perhaps for the PPC architecture although this is not clear, is in page 14 ( When Processors Hit the Power Wall (or "When the CPU hits the fan") ).
EDIT: just saw in the other thread that the CELL-64b Power link is indeed what you mean. So, could the next PPC 9xx be CELL-like? I am confused at this point.
10.4 Creating the BlueGene/L Supercomputer from Low-Power SoC ASICs
A. Bright1, M. Ellavsky2, A. Gara1, R. Haring1, R. Lembach2, J. Marcella2 and V. Salapura1
1IBM, Yorktown Heights, NY
2IBM, Rochester, MN
An overview of the design aspects of the BlueGene/L chip, the heart of the BlueGene/L supercomputer, is presented. Following an SoC approach, processors, memory and communication subsystems are integrated into one low-power chip. The high-density system packaging of the BlueGene/L system provides better power and cost performance.
(emphasis added).
So...
1) IBM is working on a multi-core 64 bit Power processer with SoC.
2) IBM is using a low power processor with SoC to create supercomputers.
This being a rumour mill, it's fair to speculate that IBM's 64 bit "Cell" processor being discussed at 9:00 A.M. is the same processor that IBM is using to build a top shelf super-computer. This would be pretty massive given that cell processors have been largely associated with video game consoles.
If IBM's supercomputer is based upon low-power processors, (i.e., 7-25 watts) which are designed from the ground up to be strung together... then THAT is a chip which is relevant to desktop computing. A chip like could see duty as a notebook processor, coupled for a desktop, a tetrad for a workstation, and 5000 for IBM's supercomputers.
EDIT: just saw in the other thread that the CELL-64b Power link is indeed what you mean. So, could the next PPC 9xx be CELL-like? I am confused at this point.
I don't know anything about Apple's plans. I do know that they are using Power processors in their machines, and these are designed and built by IBM. If IBM creates something their business is to sell it. Lots of it.
At the same time we have a Sony exec showing up on stage at MacWorld. We also have Apple creating CoreAudio, CoreImage, etc. -- all perfectly suited to hiding computational hardware details for the poor application programmer. There was even talk a couple of years ago about an Apple machine with vector units in it.
At some point 1+1 is going to equal 2.
That blurb from the program is also interesting in that it mentions on-chip memory controller and I/O. I'm surprised nobody looked at that.
And just one more thing before I submit: 5 years ago IBM published a paper that discussed "the cellular architecture". Everyone assumes that Sony and Toshiba have lots of IP tied up in the Cell. IBM has been building high performance computers for a long time, and they have zillions of patents and tons of expertise. What if Sony and Toshiba were just providing funding and some performance/power targets, plus buying a manufacturing license? That changes the picture quite a bit, doesn't it?
Comments
Sit back and watch steve-o work the car analogies all over again...
i'm really looking forward to the WWDC this year.
Originally posted by Krassy
why in 2007? it's way too long. i want this thing with 2.5 GHz now at 90nm - and i don't think that it should be a real problem for IBM to manufacture something like that... ?
what about graphics? X800 ?
Why are people in these forums so anxious to get a G6? The G5 is a young chip...one reason I want an apple is because they don't become obsolete as fast, but the way people talk on these forums its like as soon as one thing comes out, they want an upgrade. I wouldn't hope to see a G6 until 2007, because I'd like to see them milk the G5 for all its worth, before getting a new chip and building a new OS around it. I plan to get a G5 iMac or eMac(depending on whether or not they upgrade it) at the end of august, once I sell my current PC, and I'd hate for a G6 to be announced right after a buy it. There's something to be said of longevity, and thats why Apple appeals to me.
- your g5 imac would be outdated
- apple will build a new os around that chip (os x is here to stay for a very long time)
- it would be a bad thing for the pro-market
- i wouldn't be happy with such a machine
- my G4/400 won't run the current OS good enough to be a cool computer
Originally posted by Alcibiades
Why are people in these forums so anxious to get a G6? The G5 is a young chip...one reason I want an apple is because they don't become obsolete as fast, but the way people talk on these forums its like as soon as one thing comes out, they want an upgrade. I wouldn't hope to see a G6 until 2007, because I'd like to see them milk the G5 for all its worth, before getting a new chip and building a new OS around it. I plan to get a G5 iMac or eMac(depending on whether or not they upgrade it) at the end of august, once I sell my current PC, and I'd hate for a G6 to be announced right after a buy it. There's something to be said of longevity, and thats why Apple appeals to me.
Now, things are really getting interesting. Technologies and designs that had been confined to workstations are migrating down to PCs (and farther down, to consoles and other dedicated devices) in earnest. This began a few years ago with the adoption of HyperTransport, and it's accelerating, courtesy of AMD, NVIDIA and IBM especially.
Previously, there was a great deal of pressure on the CPU in PCs simply because high-speed connections elsewhere on the board were expensive, so the CPU had to do everything (the Amiga stands out as an early and noteworthy exception). Now, high-speed connections between chips on the motherboard make it more feasible to farm work out. The massive inequality in clockspeeds between computational units and memory (which used to go the other way, hard as that may be to believe) has made it possible to do some "precalculation" work on data as it streams by, before it gets to the CPU, or the GPU, or wherever the intended destination is (IBM attempted this early in PCs with their Microchannel architecture—but the less said about that, the better).
Dual cores are now feasible; so are multithreaded architectures. So are clusters of independent, loosely coupled processors on a single die (Cell, essentially). The combination of mainstream multitasking and multithreading operating systems (even in handhelds!), an inflection point in the number of transistors available at the right cost, and some other factors that I'm probably missing (well, the constant demands from all sides that will assault a CPU from the various components of a motherboard with high-bandwidth interconnects!), coupled with the most recent obstacles encountered at 90nm and below, lead me to think that we'll be seeing some really ingenious CPU designs in the next few years, popping up in surprising places.
And I can't wait to see them.
Originally posted by Amorph
And I can't wait to see them.
Pay close attention next week, then.
Originally posted by Programmer
Pay close attention next week, then.
Vewy intewesting.
At first I was like
But then I thought
I mean, "announcements" can precede release by months, if not years.
Of course, that made me
And if Programmer was just jerking us around I'd be
After awhile I calmed down and was
But I finally saw the absuridty of the whole thing. So I
Originally posted by g3pro
I'm still waiting for the quad-core G4.
The interesting thing is that this isn't completly out of the question as a future Freescale product if you look at what the other embedded CPU manufacturers are doing. Not saying that it will/must happen, but it's not out of the question.
Originally posted by Programmer
Pay close attention next week, then.
Anything in particular we should pay close attention to? It is a rather broad industry, after all. I'm not asking for a hint, just a general direction.
Originally posted by cubist
Anything in particular we should pay close attention to? It is a rather broad industry, after all.
Maybe he makes reference to the ISSCC, taking place next week in San Francisco. Here is the program. i don't see however why this would be relevant to the Macintosh platform.
Originally posted by Programmer
Pay close attention next week, then.
Like PB, I think Programmer is referring to the ISSCC (IEEE International Solid-State Circuits Conference) that is coming up next week.
Perhaps IBM will have more to talk about than the "Cell" chip.
Originally posted by PB
i don't see however why this would be relevant to the Macintosh platform.
It might not be; but Programmer was answering my wish to see ingenious new architectures. I don't really care where they end up, I just want to see them.
It's more or less a given that if Apple has anything to show in the next few months (say, WWDC), they've asked the parties responsible for the work to please keep a lid on it.
Originally posted by Aphelion
Like PB, I think Programmer is referring to the ISSCC (IEEE International Solid-State Circuits Conference) that is coming up next week.
Perhaps IBM will have more to talk about than the "Cell" chip.
The PPC970 is a "Power" processor. Search this:
http://www.isscc.org/isscc/2005/ap/I...nceProgram.pdf
For references to Power architecture processors.
Originally posted by Programmer
For references to Power architecture processors.
The only reference, and this could remotely only be relevant, is in page 36:
A CELL Processor is a multi-core chip consisting of a 64b Power architecture processor, multiple streaming processors, a flexible IO interface, and a memory interface controller.This SoC is implemented in 90nm SOI technology. The chip is designed with a high degree of modularity and reuse to maximize the custom circuit content and achieve a high-frequency clock-rate.
Another topic, of interest perhaps for the PPC architecture although this is not clear, is in page 14 ( When Processors Hit the Power Wall (or "When the CPU hits the fan") ).
EDIT: just saw in the other thread that the CELL-64b Power link is indeed what you mean. So, could the next PPC 9xx be CELL-like? I am confused at this point.
10.4 Creating the BlueGene/L Supercomputer from Low-Power SoC ASICs
A. Bright1, M. Ellavsky2, A. Gara1, R. Haring1, R. Lembach2, J. Marcella2 and V. Salapura1
1IBM, Yorktown Heights, NY
2IBM, Rochester, MN
An overview of the design aspects of the BlueGene/L chip, the heart of the BlueGene/L supercomputer, is presented. Following an SoC approach, processors, memory and communication subsystems are integrated into one low-power chip. The high-density system packaging of the BlueGene/L system provides better power and cost performance.
(emphasis added).
So...
1) IBM is working on a multi-core 64 bit Power processer with SoC.
2) IBM is using a low power processor with SoC to create supercomputers.
This being a rumour mill, it's fair to speculate that IBM's 64 bit "Cell" processor being discussed at 9:00 A.M. is the same processor that IBM is using to build a top shelf super-computer. This would be pretty massive given that cell processors have been largely associated with video game consoles.
If IBM's supercomputer is based upon low-power processors, (i.e., 7-25 watts) which are designed from the ground up to be strung together... then THAT is a chip which is relevant to desktop computing. A chip like could see duty as a notebook processor, coupled for a desktop, a tetrad for a workstation, and 5000 for IBM's supercomputers.
Originally posted by PB
EDIT: just saw in the other thread that the CELL-64b Power link is indeed what you mean. So, could the next PPC 9xx be CELL-like? I am confused at this point.
I don't know anything about Apple's plans. I do know that they are using Power processors in their machines, and these are designed and built by IBM. If IBM creates something their business is to sell it. Lots of it.
At the same time we have a Sony exec showing up on stage at MacWorld. We also have Apple creating CoreAudio, CoreImage, etc. -- all perfectly suited to hiding computational hardware details for the poor application programmer. There was even talk a couple of years ago about an Apple machine with vector units in it.
At some point 1+1 is going to equal 2.
That blurb from the program is also interesting in that it mentions on-chip memory controller and I/O. I'm surprised nobody looked at that.
And just one more thing before I submit: 5 years ago IBM published a paper that discussed "the cellular architecture". Everyone assumes that Sony and Toshiba have lots of IP tied up in the Cell. IBM has been building high performance computers for a long time, and they have zillions of patents and tons of expertise. What if Sony and Toshiba were just providing funding and some performance/power targets, plus buying a manufacturing license? That changes the picture quite a bit, doesn't it?
Originally posted by Programmer
The PPC970 is a "Power" processor. Search this:
http://www.isscc.org/isscc/2005/ap/I...nceProgram.pdf
For references to Power architecture processors.
Easy enough. There are mentions of the Cell processor on pages 29, 36 and 73. Looks like some better information on Cell is about to be published.
(edit) Sorry, took too long seaching, and you guys beat me to it. Suppose I should refresh AI while I'm looking at other things...