<strong>yah think? for a couple of million chips a year, i could see ibm shutting up (for a little bit)
and maybe this little bit is until oct. and not jan feb march....</strong><hr></blockquote>
- If IBM doesn´t have a strategic advantage in announsing the chip long time before its introduced (like saying "hey don´t buy servers from Dell right now. We have a new fantastic server chips around the corner")
- If IBM doesn´t feel it nessesary to brag about their new lab factory by saying "hey look what kind of fantastic chip we will be able to produce here in a half year" (afaik the factory is to used for producing chips for other companies, no?)
Then maybe IBM will take Apples needs into consideration .
I´m glad if all this about the next generation of chips is going to be made by IBM turns out to be true. But it also have its disadvantages. First Apple won´t have two chip makers to play out against eachother and second IBM is not going to be more considerous towards Apple than Motorola. If supplying Apple with chips is such a good buisness Motorola would have had much more focus on that. Its only because IBM will have other advantages with this new chip design that Apple is on the boat.
I´m not saying that the chip is not just around the corner. I hope it is. But don´t take the time IBM is announcing it as any sign of anything.
<strong>Aphelion, i agree. the $1,000,000 question is:
why a new plant for this type of chip?
soley because ibm is going to be more agressive with linux and their current servers?
too early for the playstation 3 for it to be it?
or apple is going to be a big purchaser and they need the capcity.
or all of the above or someother thing we have no clue of.</strong><hr></blockquote>
[quote]from the Oct 2000 Press Release:
<strong>The world of e-business is driving a massive build-out of the infrastructure of computing and communications," said Lou Gerstner, IBM Chairman and CEO. "That, in turn, drives demand for critical technical components like chips. Demand is white-hot in three critical segments -- chips for big servers, chips to power the explosion in Internet access devices and chips in the networking equipment that ties everything together. That's why today's announcement is important -- important for our industry, our customers and our employees</strong><hr></blockquote>
Well the new IBM fab was proposed and approved to handle forecasted demand in the above stated areas, of course things have changed a bit since then. Apple and IBM may be bailing each other out on this one, given that the minimal information released recently points to PowerPC chips in desktops and entry level servers. (can you say SuperMacs?).
IBM needs volume, and fast, for this fab to pay off (ROI), With the PS-3 still aways off Apple is the only manufacturer that can use these chips NOW. Snoopy had it very close, a high level Workstation for all the new apps is coming soon. Stay tuned for upheaval in the New Market.
A couple points about IBM and a 'Desktop' PowerPC.
1. They tried targetting PowerPC at the mainstream Intel desktop market once already in the mid-90s, and after immense hype, it was a complete and total failure. They probably didn't sell more than a couple thousand machines -- absolutely nobody wanted them. And PowerPC was much stronger relative to Intel in those days.
2. Since introduction of the PC, all of IBM's attempts at marketing to the end user have been total failures (see OS/2). The 'New IBM' is smart and sticks to what it's good at -- server technology and services. They've invested a bunch in Linux, but 0 of it is in desktop space.
3. The big sell of Linux in server space is because it allows companies to dump 'proprietary' hardware and replace it with cheap 'industry standard' (x86) gear (or, in IBM's case to extend their proprietary OSes with standards-based tech). Nobody's going to switch to Linux on a platform percieved to be non-standard -- they'll just stick with Sun etc.
4. Someone said that the x86 world will have trouble going 64-bit. We'll see if that's true when AMD ships x86-64 Hammer chips Q103. My guess is that you'll see IBM-branded Hammer desktops before you see anything 64-bit and PowerPC in that space.
5. PowerPC is really an industry lame-duck. Nobody (except Apple customers) is saying "Ooooo. I want that!". Even a sexy new Power4 workstation chip is not going to change the fact that it's basically a 1 customer CPU, and that's not tenable in the long run (meaning Apple's going to have to 'switch' eventually).
6. The only thing IBM really needs is a handful of Power4 workstation machines to sell to AIX developers. This market is so tiny, IBM could almost tell them to go buy Apple hardware
IMO, if you had any credibility you just blew it. I can't see anyone from IBM referring to a new chip as the "G5". "We've a new chip", maybe, but G5, no way.</strong><hr></blockquote>
Alrighty... I was a little surprised to hear them say 'G5'. I'm not sure if they used that term for my benifit, or if it the term that IBM folks know Apple is going to call it. It was the term that the IBM person used.
<strong>A couple points about IBM and a 'Desktop' PowerPC.
1. They tried targetting PowerPC at the mainstream Intel desktop market once already in the mid-90s, and after immense hype, it was a complete and total failure. They probably didn't sell more than a couple thousand machines -- absolutely nobody wanted them. And PowerPC was much stronger relative to Intel in those days.</strong><hr></blockquote>
True, but he CHRP didn't fail on it's own merits. it failed because a small software company called MS decided to not support the platfrom.
[quote]<strong>2. Since introduction of the PC, all of IBM's attempts at marketing to the end user have been total failures (see OS/2). The 'New IBM' is smart and sticks to what it's good at -- server technology and services. They've invested a bunch in Linux, but 0 of it is in desktop space.</strong><hr></blockquote>
You don't know that. They've had this 64 bit processor in development for about 3 yrs now, how do you kknow they haven't had a parallel Linux development effort also?
[quote]<strong>3. The big sell of Linux in server space is because it allows companies to dump 'proprietary' hardware and replace it with cheap 'industry standard' (x86) gear (or, in IBM's case to extend their proprietary OSes with standards-based tech). Nobody's going to switch to Linux on a platform percieved to be non-standard -- they'll just stick with Sun etc.
4. Someone said that the x86 world will have trouble going 64-bit. We'll see if that's true when AMD ships x86-64 Hammer chips Q103. My guess is that you'll see IBM-branded Hammer desktops before you see anything 64-bit and PowerPC in that space.</strong><hr></blockquote>
Perhaps. But now people will have real choice in the 64-bit arena. That can't help but be a good thing, right?
[quote]<strong>5. PowerPC is really an industry lame-duck. Nobody (except Apple customers) is saying "Ooooo. I want that!". Even a sexy new Power4 workstation chip is not going to change the fact that it's basically a 1 customer CPU, and that's not tenable in the long run (meaning Apple's going to have to 'switch' eventually).</strong><hr></blockquote>
Now here's where you totally miss the mark.
First of all, the PowerPC family is comprised of several chip architectures that all support the PPC ISA. PPC is by no means a "lame-duck", "1 custormer CPU' since it is actually a very robust ISA with strong products in both the workstation and server markets.
And it's quite "feasible" in the long run since it has the advantage of supporting very power-efficient chip designs. (Why do you think they named it "Power" PC?)
[quote]<strong>6. The only thing IBM really needs is a handful of Power4 workstation machines to sell to AIX developers. This market is so tiny, IBM could almost tell them to go buy Apple hardware </strong><hr></blockquote>
Unless, of course, this chip (and the chips that follow) spank Intel and AMD products so badly that OEMs like Dell and HP pressure MS to support the PowerPC ISA in order to keep from losing marketshare to IBM and Apple. At which point we could see the PC world switching to PPC -- a supposition no less absurd than Apple marketing an x86 ISA box.
[quote]True, but he CHRP didn't fail on it's own merits. it failed because a small software company called MS decided to not support the platfrom.<hr></blockquote>
I don't know about "it's own merits". But nobody wanted to buy the thing and therefore neither Moto, IBM, or MS wanted to continue to pay for development of WinNT/PPC (and OS/2-PPC), and that's why it went away. Even so, I don't think NT/PPC was actually cancelled until Moto pulled the last hardware off the market.
Understand that it's in MS's best interest to have strong platform competitors to Intel. They did invest quite a bit in Alpha (also without much effect).
Since we have historical evidence that NT/RISC is just not appealing to the mainstream PC market, someone needs to put forward a real argument would be different this time. (Other than "it will be faster", which is really debatable considering the last several years of PPC products and Intel/AMD's roadmaps.)
[quote]"1 custormer CPU" <hr></blockquote>
If there's someone else in the market for desktop PPC chips, I think we'd all like to hear about it.
The whole reason Apple got board with PowerPC in the first place is because they thought it would have broad industry support which would help them in their position on the thin edge of the nitch. Unfortunately, itjust didn't happen, at least in the computer market. The problems with Motorola are basically that (to them) Macs are a second tier application behind their embedded bread-and-butter (which neither needs nor wants 2Ghz CPUs).
[quote]Unless, of course, this chip (and the chips that follow) spank Intel and AMD products so badly that OEMs like Dell and HP pressure MS to support the PowerPC ISA in order to keep from losing marketshare to IBM and Apple.<hr></blockquote>
I haven't heard talk like this since 1993 Hopefully the chip really is *that* good.
... If there's someone else in the market for desktop PPC chips, I think we'd all like to hear about it.
</strong><hr></blockquote>
Can it be so bad having a one customer CPU? IBM is making a special for Ninetendo only. The development cost of the chip is spread over the production run. If you are a high volume customer, the chips cost a little less. Also, think about OS X. That is a one customer OS. Yes, Apple is unique. The name of the game will be getting market share up where one customer has less overhead to pay for in each product shipped.
The question about CPU should be performance for the market. I believe Apple is going after workstations for the film industry and wants to take it over. I think an IBM 64-bit PPC gives Apple the surperior performance a workstation needs to run Shake, and Apple has a good chance of walking away with the market.
Can it be so bad having a one customer CPU? IBM is making a special for Ninetendo only. The development cost of the chip is spread over the production run. [QB]<hr></blockquote>
My understanding is that the Nintendo chips is just a slightly modded G3 750.
Intel can spread P4 R&D costs over 60% of the market. AMD has about 20%. Apple has 3 or 4% marketshare and those users expect the same results as Intel/AMD.
Now, IBM can ship $2000 CPUs in their big servers, but there's no way Apple can do that up against $500 Xeons and Hammers and hit their margins. Sumthin has to give.
Note that when PowerPC was introduced, Apple had about a 15% marketshare, and AIM was predicting getting about 10-25% of the Windows market. Big difference between 25% and 4%.
<strong>Usually means production volumes somewhere around 6 mos from samples unless something goes wrong.
So, it's possible we could have production quantities of Macs using this chip (which assumes it is a Mac chip) by MWSF if the samples live up to expectations and they don't have to do major revisions to the design.
Of course if you (or your little bird) meant next March or April, then we won't see them 'til next summer, if then.</strong><hr></blockquote>
no, definitely not. i was told at the beginning of the year that ibm's chip would be sampling in a couple months. so i just esimated the timing to be more specific.
Now, IBM can ship $2000 CPUs in their big servers, but there's no way Apple can do that up against $500 Xeons and Hammers and hit their margins. Sumthin has to give.</strong><hr></blockquote>
What gives, so to speak, is that for design reasons the PPCs are cheaper to design, and also to fab.
If this new IBM processor is a slimmed-down Power4 then most of the heavy R&D work is already done.
Isn't the Power 4 a processor that is already available? If so isn' it just a case of the altivec unit being added to the general design. If so I hardly see how it could be months away from production.
<strong>Isn't the Power 4 a processor that is already available? If so isn' it just a case of the altivec unit being added to the general design. If so I hardly see how it could be months away from production.</strong><hr></blockquote>
The POWER4 is a giant, multi-cored monster. There's no way it will appear in a Mac. The chip this thread is dwelling on is based on the POWER4, but it's significantly redesigned for use in the desktop market.
<strong>Isn't the Power 4 a processor that is already available? If so isn' it just a case of the altivec unit being added to the general design. If so I hardly see how it could be months away from production.</strong><hr></blockquote>
Very cute... <img src="graemlins/lol.gif" border="0" alt="[Laughing]" />
Ummm you were kidding right? If on the off chance you aren't well I think you have a lot of reading to do but in short this new chip is NOT the Power4 (used in very expensive IBM server boxes) but it is BASED ON the Power4.
This CPU has been built with the 'desktop' and 'small server' in mind. That means tons of stuff will have to change to get the cost of CPU production down to keeping the heat down to keeping power usage down etc etc etc. Also you speak of adding altivec as if were as simple as adding jelly to a peanut-butter sandwich.
<strong>I´m glad if all this about the next generation of chips is going to be made by IBM turns out to be true. But it also have its disadvantages. First Apple won´t have two chip makers to play out against eachother and second IBM is not going to be more considerous towards Apple than Motorola. If supplying Apple with chips is such a good buisness Motorola would have had much more focus on that. Its only because IBM will have other advantages with this new chip design that Apple is on the boat.</strong><hr></blockquote>
'two chip makers to play out against each other'
Yea well what good has it done them over the past three years? Do you think MOT would have given us less than a 500Mhz pop (err 750Mhz 'soon') over those oh-so-many-years if we didn't have IBM to use 'against' MOT? Sheesh then thank goodness we had IBM to play against MOT I say.
'IBM is not going to be more considerous towards Apple than Motorola'
IBM makes very high performance PPC cpu's used in $50k+ hardware... MOT makes cell phones and CPU's for the auto industry. I think IBM has a much better chance at giving Apple what it needs - MOT has had 3 years to show what it can do and lets face facts it couldn't do much. From where I sit a total speed boost of 500Mhz and zero signs of moving the G4 to DDR support pretty much tells the whole story.
If someone can tell me ANYTHING that MOT has done in the past three years that you could say was even mildly impressive I'd love to hear it. (I give full credit for AltiVec but that was over three years ago - since then MOT has died a slow death in the desktop CPU space).
We have nothing to loose with the move to IBM and if things turn out even half as good as they sound right now we have everything to gain.
Comments
<strong>yah think? for a couple of million chips a year, i could see ibm shutting up (for a little bit)
and maybe this little bit is until oct. and not jan feb march....</strong><hr></blockquote>
- If IBM doesn´t have a strategic advantage in announsing the chip long time before its introduced (like saying "hey don´t buy servers from Dell right now. We have a new fantastic server chips around the corner")
- If IBM doesn´t feel it nessesary to brag about their new lab factory by saying "hey look what kind of fantastic chip we will be able to produce here in a half year" (afaik the factory is to used for producing chips for other companies, no?)
Then maybe IBM will take Apples needs into consideration .
I´m glad if all this about the next generation of chips is going to be made by IBM turns out to be true. But it also have its disadvantages. First Apple won´t have two chip makers to play out against eachother and second IBM is not going to be more considerous towards Apple than Motorola. If supplying Apple with chips is such a good buisness Motorola would have had much more focus on that. Its only because IBM will have other advantages with this new chip design that Apple is on the boat.
I´m not saying that the chip is not just around the corner. I hope it is. But don´t take the time IBM is announcing it as any sign of anything.
<strong>Aphelion, i agree. the $1,000,000 question is:
why a new plant for this type of chip?
soley because ibm is going to be more agressive with linux and their current servers?
too early for the playstation 3 for it to be it?
or apple is going to be a big purchaser and they need the capcity.
or all of the above or someother thing we have no clue of.</strong><hr></blockquote>
[quote]from the Oct 2000 Press Release:
<strong>The world of e-business is driving a massive build-out of the infrastructure of computing and communications," said Lou Gerstner, IBM Chairman and CEO. "That, in turn, drives demand for critical technical components like chips. Demand is white-hot in three critical segments -- chips for big servers, chips to power the explosion in Internet access devices and chips in the networking equipment that ties everything together. That's why today's announcement is important -- important for our industry, our customers and our employees</strong><hr></blockquote>
Well the new IBM fab was proposed and approved to handle forecasted demand in the above stated areas, of course things have changed a bit since then. Apple and IBM may be bailing each other out on this one, given that the minimal information released recently points to PowerPC chips in desktops and entry level servers. (can you say SuperMacs?).
IBM needs volume, and fast, for this fab to pay off (ROI), With the PS-3 still aways off Apple is the only manufacturer that can use these chips NOW. Snoopy had it very close, a high level Workstation for all the new apps is coming soon. Stay tuned for upheaval in the New Market.
[ 08-24-2002: Message edited by: Aphelion ]</p>
1. They tried targetting PowerPC at the mainstream Intel desktop market once already in the mid-90s, and after immense hype, it was a complete and total failure. They probably didn't sell more than a couple thousand machines -- absolutely nobody wanted them. And PowerPC was much stronger relative to Intel in those days.
2. Since introduction of the PC, all of IBM's attempts at marketing to the end user have been total failures (see OS/2). The 'New IBM' is smart and sticks to what it's good at -- server technology and services. They've invested a bunch in Linux, but 0 of it is in desktop space.
3. The big sell of Linux in server space is because it allows companies to dump 'proprietary' hardware and replace it with cheap 'industry standard' (x86) gear (or, in IBM's case to extend their proprietary OSes with standards-based tech). Nobody's going to switch to Linux on a platform percieved to be non-standard -- they'll just stick with Sun etc.
4. Someone said that the x86 world will have trouble going 64-bit. We'll see if that's true when AMD ships x86-64 Hammer chips Q103. My guess is that you'll see IBM-branded Hammer desktops before you see anything 64-bit and PowerPC in that space.
5. PowerPC is really an industry lame-duck. Nobody (except Apple customers) is saying "Ooooo. I want that!". Even a sexy new Power4 workstation chip is not going to change the fact that it's basically a 1 customer CPU, and that's not tenable in the long run (meaning Apple's going to have to 'switch' eventually).
6. The only thing IBM really needs is a handful of Power4 workstation machines to sell to AIX developers. This market is so tiny, IBM could almost tell them to go buy Apple hardware
<strong>
IMO, if you had any credibility you just blew it. I can't see anyone from IBM referring to a new chip as the "G5". "We've a new chip", maybe, but G5, no way.</strong><hr></blockquote>
Alrighty... I was a little surprised to hear them say 'G5'. I'm not sure if they used that term for my benifit, or if it the term that IBM folks know Apple is going to call it. It was the term that the IBM person used.
<strong>A couple points about IBM and a 'Desktop' PowerPC.
1. They tried targetting PowerPC at the mainstream Intel desktop market once already in the mid-90s, and after immense hype, it was a complete and total failure. They probably didn't sell more than a couple thousand machines -- absolutely nobody wanted them. And PowerPC was much stronger relative to Intel in those days.</strong><hr></blockquote>
True, but he CHRP didn't fail on it's own merits. it failed because a small software company called MS decided to not support the platfrom.
[quote]<strong>2. Since introduction of the PC, all of IBM's attempts at marketing to the end user have been total failures (see OS/2). The 'New IBM' is smart and sticks to what it's good at -- server technology and services. They've invested a bunch in Linux, but 0 of it is in desktop space.</strong><hr></blockquote>
You don't know that. They've had this 64 bit processor in development for about 3 yrs now, how do you kknow they haven't had a parallel Linux development effort also?
[quote]<strong>3. The big sell of Linux in server space is because it allows companies to dump 'proprietary' hardware and replace it with cheap 'industry standard' (x86) gear (or, in IBM's case to extend their proprietary OSes with standards-based tech). Nobody's going to switch to Linux on a platform percieved to be non-standard -- they'll just stick with Sun etc.
4. Someone said that the x86 world will have trouble going 64-bit. We'll see if that's true when AMD ships x86-64 Hammer chips Q103. My guess is that you'll see IBM-branded Hammer desktops before you see anything 64-bit and PowerPC in that space.</strong><hr></blockquote>
Perhaps. But now people will have real choice in the 64-bit arena. That can't help but be a good thing, right?
[quote]<strong>5. PowerPC is really an industry lame-duck. Nobody (except Apple customers) is saying "Ooooo. I want that!". Even a sexy new Power4 workstation chip is not going to change the fact that it's basically a 1 customer CPU, and that's not tenable in the long run (meaning Apple's going to have to 'switch' eventually).</strong><hr></blockquote>
Now here's where you totally miss the mark.
First of all, the PowerPC family is comprised of several chip architectures that all support the PPC ISA. PPC is by no means a "lame-duck", "1 custormer CPU' since it is actually a very robust ISA with strong products in both the workstation and server markets.
And it's quite "feasible" in the long run since it has the advantage of supporting very power-efficient chip designs. (Why do you think they named it "Power" PC?)
[quote]<strong>6. The only thing IBM really needs is a handful of Power4 workstation machines to sell to AIX developers. This market is so tiny, IBM could almost tell them to go buy Apple hardware
Unless, of course, this chip (and the chips that follow) spank Intel and AMD products so badly that OEMs like Dell and HP pressure MS to support the PowerPC ISA in order to keep from losing marketshare to IBM and Apple. At which point we could see the PC world switching to PPC -- a supposition no less absurd than Apple marketing an x86 ISA box.
I don't know about "it's own merits". But nobody wanted to buy the thing and therefore neither Moto, IBM, or MS wanted to continue to pay for development of WinNT/PPC (and OS/2-PPC), and that's why it went away. Even so, I don't think NT/PPC was actually cancelled until Moto pulled the last hardware off the market.
Understand that it's in MS's best interest to have strong platform competitors to Intel. They did invest quite a bit in Alpha (also without much effect).
Since we have historical evidence that NT/RISC is just not appealing to the mainstream PC market, someone needs to put forward a real argument would be different this time. (Other than "it will be faster", which is really debatable considering the last several years of PPC products and Intel/AMD's roadmaps.)
[quote]"1 custormer CPU" <hr></blockquote>
If there's someone else in the market for desktop PPC chips, I think we'd all like to hear about it.
The whole reason Apple got board with PowerPC in the first place is because they thought it would have broad industry support which would help them in their position on the thin edge of the nitch. Unfortunately, itjust didn't happen, at least in the computer market. The problems with Motorola are basically that (to them) Macs are a second tier application behind their embedded bread-and-butter (which neither needs nor wants 2Ghz CPUs).
[quote]Unless, of course, this chip (and the chips that follow) spank Intel and AMD products so badly that OEMs like Dell and HP pressure MS to support the PowerPC ISA in order to keep from losing marketshare to IBM and Apple.<hr></blockquote>
I haven't heard talk like this since 1993
[ 08-26-2002: Message edited by: IntlHarvester ]</p>
<strong>
... If there's someone else in the market for desktop PPC chips, I think we'd all like to hear about it.
</strong><hr></blockquote>
Can it be so bad having a one customer CPU? IBM is making a special for Ninetendo only. The development cost of the chip is spread over the production run. If you are a high volume customer, the chips cost a little less. Also, think about OS X. That is a one customer OS. Yes, Apple is unique. The name of the game will be getting market share up where one customer has less overhead to pay for in each product shipped.
The question about CPU should be performance for the market. I believe Apple is going after workstations for the film industry and wants to take it over. I think an IBM 64-bit PPC gives Apple the surperior performance a workstation needs to run Shake, and Apple has a good chance of walking away with the market.
[QB]
Can it be so bad having a one customer CPU? IBM is making a special for Ninetendo only. The development cost of the chip is spread over the production run. [QB]<hr></blockquote>
My understanding is that the Nintendo chips is just a slightly modded G3 750.
Intel can spread P4 R&D costs over 60% of the market. AMD has about 20%. Apple has 3 or 4% marketshare and those users expect the same results as Intel/AMD.
Now, IBM can ship $2000 CPUs in their big servers, but there's no way Apple can do that up against $500 Xeons and Hammers and hit their margins. Sumthin has to give.
Note that when PowerPC was introduced, Apple had about a 15% marketshare, and AIM was predicting getting about 10-25% of the Windows market. Big difference between 25% and 4%.
<strong>Usually means production volumes somewhere around 6 mos from samples unless something goes wrong.
So, it's possible we could have production quantities of Macs using this chip (which assumes it is a Mac chip) by MWSF if the samples live up to expectations and they don't have to do major revisions to the design.
Of course if you (or your little bird) meant next March or April, then we won't see them 'til next summer, if then.</strong><hr></blockquote>
no, definitely not. i was told at the beginning of the year that ibm's chip would be sampling in a couple months. so i just esimated the timing to be more specific.
<strong>
Now, IBM can ship $2000 CPUs in their big servers, but there's no way Apple can do that up against $500 Xeons and Hammers and hit their margins. Sumthin has to give.</strong><hr></blockquote>
What gives, so to speak, is that for design reasons the PPCs are cheaper to design, and also to fab.
If this new IBM processor is a slimmed-down Power4 then most of the heavy R&D work is already done.
<strong>Isn't the Power 4 a processor that is already available? If so isn' it just a case of the altivec unit being added to the general design. If so I hardly see how it could be months away from production.</strong><hr></blockquote>
The POWER4 is a giant, multi-cored monster. There's no way it will appear in a Mac. The chip this thread is dwelling on is based on the POWER4, but it's significantly redesigned for use in the desktop market.
<strong>Isn't the Power 4 a processor that is already available? If so isn' it just a case of the altivec unit being added to the general design. If so I hardly see how it could be months away from production.</strong><hr></blockquote>
Very cute... <img src="graemlins/lol.gif" border="0" alt="[Laughing]" />
Ummm you were kidding right? If on the off chance you aren't well I think you have a lot of reading to do but in short this new chip is NOT the Power4 (used in very expensive IBM server boxes) but it is BASED ON the Power4.
This CPU has been built with the 'desktop' and 'small server' in mind. That means tons of stuff will have to change to get the cost of CPU production down to keeping the heat down to keeping power usage down etc etc etc. Also you speak of adding altivec as if were as simple as adding jelly to a peanut-butter sandwich.
It ain't...
Dave
First Apple made a deal with the then-devil (MS) and now it appears like Apple is selling it soul to the old devil.
Be careful Apple. He might return the hammer...
<strong>I´m glad if all this about the next generation of chips is going to be made by IBM turns out to be true. But it also have its disadvantages. First Apple won´t have two chip makers to play out against eachother and second IBM is not going to be more considerous towards Apple than Motorola. If supplying Apple with chips is such a good buisness Motorola would have had much more focus on that. Its only because IBM will have other advantages with this new chip design that Apple is on the boat.</strong><hr></blockquote>
'two chip makers to play out against each other'
Yea well what good has it done them over the past three years? Do you think MOT would have given us less than a 500Mhz pop (err 750Mhz 'soon') over those oh-so-many-years if we didn't have IBM to use 'against' MOT? Sheesh then thank goodness we had IBM to play against MOT I say.
'IBM is not going to be more considerous towards Apple than Motorola'
IBM makes very high performance PPC cpu's used in $50k+ hardware... MOT makes cell phones and CPU's for the auto industry. I think IBM has a much better chance at giving Apple what it needs - MOT has had 3 years to show what it can do and lets face facts it couldn't do much. From where I sit a total speed boost of 500Mhz and zero signs of moving the G4 to DDR support pretty much tells the whole story.
If someone can tell me ANYTHING that MOT has done in the past three years that you could say was even mildly impressive I'd love to hear it. (I give full credit for AltiVec but that was over three years ago - since then MOT has died a slow death in the desktop CPU space).
We have nothing to loose with the move to IBM and if things turn out even half as good as they sound right now we have everything to gain.
Dave
[ 08-28-2002: Message edited by: DaveGee ]</p>