I can't believe that Steve has that much influence over IBM. It is very surprising that Apple was able to get IBM to remove this information. Especially since Apple has not announced it is even using this chip. To me this is even more evidence that Apple will be using it. IBM doesn't have anywhere near the concern that Apple does for preannouncing products. Only Apple would care that much to have the information removed.</strong><hr></blockquote>
This strongly suggests that Apple has subsidised the development of this chip and has legally reserved certain rights about the publicity surrounding it. It may be just a case of the left hand not knowing what the right hand was doing - this was the IBM Germany site, not USA. They may not have been as on-the-ball as they might have been, and posted info that technically trampled on Apple's rights without realizing it. Or, it may be as simple as that they didn't get clearance from Apple before posting it - nothing inherently wrong with the info, they just didn't go through the right channels and get all the signatures they needed.
I'm beginning to suspect that the 970 was actually Apple's idea in the first place. They probably pitched it to Big Blue as one they could both use, and offered to fund much of its development. Curiouser and curiouser...
[quote] I guess this thread may as well turn into an emulation/OS X x86 discussion afterall because the subject of our conversation disappeared, as Amorph noted. That's really scary indeed. One hopes that Apple simply requested the press release be pulled so it wouldn't steal the thunder. Otherwise, it would be terrible to get psyched for a 2.5GHz 970, only to be greeted by a top end of 1.8. <hr></blockquote>
Yeah, there is so much optimism going around that if reality doesn't match up with all of our hopes, there is sure to be much dissapointment, and maybe anger in some (eg "I'm switching to Windows because the new machines are ONLY 1.8Ghz! That was supposed to be the LOW end!"). We're all being built up by press releases and 970 discussion in the forums. The higher you get, the harder you hit when you fall. I can't help but think about that alongside all of the excitement.
I think hardcore gamers who aren't completely attached to Windows would switch for the performance, if the games were there. But I'm not so sure that they will be. Remember we still only have %3.
I don't think it's fair to say that the people who have moved to Windows moved simply because of performance. And I don't think its fair to say that everyone would suddenly switch when Apple's machines are the ones doing the ass kicking. Yes some people would, because they rely on high performance to get their work done to make the money to put the food on the table, and I'm not denying it. But plenty of people use computers for every day things and will be content with slower and cheaper PCs.
I think we're talking past each other a bit here. Let me try one more time, then I'll drop the subject.
I'm basing my opinion on two assumptions:
1) That the 970 offers superior enough performance that it can offer an emulated x86 environment as good as native x86 hardware - no significant performance penalty.
2) That there is a large number of Windows users who would be willing to switch to OS X were it not for their large investment in Windows software. This number would be enough to boost Apple's marketshare to the 10% range or higher.
My scenario then is that Apple offers a good IA-32 emulator within OS X. These potential switchers go to buy new hardware, and realise that they can get a Mac system that is as good as any x86 box to run their Windows software, and gets them OS X to boot. So they buy a Mac and load all their old Windows software onto it. Now, as the need arises for new versions of their software, what are they going to want? Another Windows version to run under emulation, or a native OS X version that will run significantly faster on their 970 system (even with crappy porting)? I don't think there's any question that developers will feel enormous pressure to get quality OS X native versions out. Those that don't respond will find their marketshare dropping as buyers head for their competitors.
I visualize the IA-32 emulation as being a lot like Classic mode. Crucially important to switchers for 6 months to a year, then gradually fading to almost complete disuse.
You don't have to agree with me, but that's the nuts and bolts of my argument. As I said at the start, I'll leave this topic alone now.
[quote]The hardware's purpose is simply to run the software without which the hardware has little utility other than as a decorative object or paper weight. <hr></blockquote>
Ask yourself this...
Can Windows run on anything other than some X86 processor-based system?
FrostyMMB writes:
[quote]So what? It's %3 of the market and they are still making the other %97 content. Quark is still around, but the Mac community is very angry with them. <hr></blockquote>
Um, how much do you want to bet that the majority of companies running Quark, are running it on a Mac and not some Windows crate. In certain markets, the PC has FAR LESS marketshare.
[quote]No, graphics (or video, or 3D) professionals most certainly would not. But the developers can save money and satisfy the large majority of Windows users. <hr></blockquote>
It isn't that simple... Besides, they should be diversifying.
[quote]The Mac marketshare is so small that it wouldn't make a huge impact on the success of the developer. Again, they are still making the large majority happy. <hr></blockquote>
Whoops! What it. You're making the same mistake again... You are mistakenly assuming that the Macs *overall* marketshare is am accurate representation of their marketshare in specific markets. Don't be silly. And don't forget, Apple is developing their X11 That too will aid in further development ;-) And then you are forgetting about all that continued chaos and uncertainty that will exist in the Windows world... It doesn't look like they will be getting 64-bit desktops for quite some time (end of decade?). However, what they *will* be getting is all that fancy DRM stuff . there seems to be an anti-Microsoft air going around lately. I think it's just gonna get worse.
On the marketshare thing you keep mentioning.. 3-4% could amount to market 40,000,000 potential users. That's nothing to ignore if your a developer. So, your 3% is rather misrepresented. If I were you, I'd think it through again ;-)
Still, I would find it rather disturbing knowing that my hard-earned $$ is being spent on apps coded by developers who are choosing the most ubiquitous (and arguably worst of the bunch) instead of a (possible) better option? And it's like saying that Goodyear will (should?) only make tires for the Toyota Carolla and Ford Escort because they hold the largest marketshare..
To quote my friend John C. Welch:
[quote]Neither is Maya, or Lightwave, or MatLab. But all of those either run, or are being ported to Mac OS X. The cost of the software is not an issue unless you are writing games. Remember, If it's a good tool, people pay for it. I don't think that Maya is going to grab 40 million users, but they'll get enough. MatLab is going to be an XFree86 application, ala IDL/Envi, so the port from the Unix versions of that application are pretty simple.
... I find it odd that an OS that is obviously on the way out, Irix, is considered a more viable option than Mac OS X for various reasons. - Jon C. Welch<hr></blockquote>
And Dave K. Every says:
[quote]We could just as readily ask then why does Microsoft do Office or Outlook for the Mac, etc... Because it is a very profitable segment for them. In fact, THOSE apps required complete rewrites and were much more major development efforts than what we are talking about here (since they used the Mac API's)... That's the job of software companies; create solutions to gain new markets. If you stop innovating, then you are stagnant and just waiting to be obviated by someone with more a clue (more vision) than you have.
When you port to other platforms, and gain new customers, they give you insights into new features or ways of thinking that help your entire market. This is one of the reasons why Microsoft keeps their Mac products; it makes their Windows versions BETTER. In fact, if I wasn't for the Mac products, I don't think most of the Windows apps would have been nearly as big. (Heck, most were Mac first). - (David K. Every) <hr></blockquote>
The Mac shouldn't be viewed as just "the other platform"
A significant point that is often overlooked is the fact that there is in fact a "crossplatform market". For instance, you cold look at it as three (or more) separate markets... The Mac market, the PC market and the crossplatform market (the Mac+PC market). That is the Mac+PC markets are markets where organizations use BOTH Mac *and* PC. So, If you support one platform (either one), you are actually missing TWO markets. 1. the other platform AND 2. the cross platform market. This indicates that ANY revenue generated from sales of supporting another platform (in this case Mac OS X/Unix) will also draw revenue from the *combined* market as well. How can you (or rather the developers) not see the potential benefits?
Historically, it's been *well* documented that it costs developers less to support the Mac than it does the PC. That means for every customer you convert to the Mac version you get a *complete upgrade* that you might not have sold (or at the very least a "competitive" upgrade, but the additional cost savings associated with supporting the Mac will provide additional revenue in the form of savings. Again, just another way to look at it. Most people and organizations only care to look at it from a single angle (narrow minded and shortsighted). Alias|Wavefront grew their Maya market 25% by deciding to support the Macintosh OS X platform. This should open some eyes. As a matter of fact it has, Mac development hasn't been better. and rumor is that DB/2 is the next major app headed in the direction of OS X.
[quote]Every company that is not paranoid about their markets, and thus stops trying to innovate (and penetrate new markets) in order to maximize profits of the now, sells their future for the present. This short-term quarterly report thinking is what has allowed not only companies but whole industries to be eaten alive.
Some of these companies do survive. They go into innovation through acquisition mode; and can sometimes acquire fast enough to tread water or break even. (Or at least slow the inevitable descent into oblivion). Most go under, or become pathetic shadows of what they once were, and shameful embarrassments compared to their potential.
So, these developers are choosing to *only* support Windows, but are laying the foundation for their own replacement. New platforms coming up are ignored, by them, but not their competition. They are entrenched in their current market and profit taking, and so are missing out on many opportunities that their competition isn't. Let's not learn the lessons of Novell, Lotus, DEC, Wang, DataGeneral, Packard Bell, and so on. Let's mimic them, because many of them were profitable, in the short term, right before they focused on the now (too long) and let the competition pass them by, and they disappeared. ... I tend to think that if they ignore the #2 OS, and the #1 UNIX OS, and the *fastest growing OS* (and UNIX OS), then they are not exactly being wise or planning for the future.
And you have to remember, ignoring that trend means more problems not only for Mac market, but also for all organizations that are mixed environments.
Remember, the tools are the platform. If they don't support my platform, then they are not supporting me (as I would like). This means that they will either drive customers to competing products (and as you know it is 10 times harder to get someone back once they've learned something else), or they will be telling me that my concerns don't matter to them (generating bad will). Neither is good business.
"Yet another platform" that is bigger than all but Windows (and has a few million seats). It seems like they should choose their platforms to reflect market trends, don't you think? - Dave K. Every <hr></blockquote>
<strong>Only Apple would care that much to have the information removed.</strong><hr></blockquote>
Not quite. IBM would like to see it pulled if it was factually inaccurate. As in, if the quoted speed ranges are not in line with expected results or yields.
Not quite. IBM would like to see it pulled if it was factually inaccurate. As in, if the quoted speed ranges are not in line with expected results or yields.</strong><hr></blockquote>
...and/or the prototype they showed had nothing to do with 970.
Waitasec... what if one of moki's other 'sandbags' (weird term) was processor performance? I mean, IBM kinda white-lied about processor speed, right? What if their 'predicted' SPEC benchmark results are way off the mark... on the side of pessimism?
What if another white lie was power consumption? This is a 0.13 process, low-k dielectric, SOI, copper process, is it not? That's what East Fishkill (what a gross name) was all about: implementing ALL of IBM's process technologies to create chips that a) perform like crazy and b) draw less power.
0.9 is next, and coming fast from the sound of it. Man, 64-bit Powerbook 3D workstations are going to roXxor.
Just a thought.
edit: No one's going to complain (except Intel, hee hee) if IBM under-promises and over-delivers, right? No SEC investigations for stock manipulation or anything.
And the low-k dielectric & Silicon On Insulator (SOI) are a two way street: either faster speed chips drawing acceptablly low power, or hella fast chips drawing crazy power (SOI & LKD redue 'cross talk' AKA electrical interference by insulating everything, increasing reliability of signals).
either way,
*crrrrunnnk* **Krunnnnkkk** **cccrrrruuunnnkkk***
What's that grinding sound?
Oh wait, that's IBM opening up a fresh can of whup-ass!
Indeed they are!!! I wonder if there will be a way to accomodate near the same RAM capacity as will be available to the desktops with the 970 (beyond 4GB) in the PowerBook -- I would think adding two more slots wouldn't be hard (but wait, 1GB chips?). I would like to see Apple investigate cooling in the laptops a little more before they do any more aesthetic leaps and bounds (it seems if that had been done before the 17" was announced lots of people would have had them now).
Yes, it's occured to me. It's also occuring to Apple that making such a big unit so thin in mass-production is a pain in the neck - hence the broken units and the problems they've been having with the 1GHz G4s. The main reason they went to it was of its lower cost and lighter weight.
Anyway, back to the 970. 1.8 GHz will be wonderful if we get it this fall. I'd go in for one alongside this TiBook. I don't think Apple'd push the 2.5GHz procs in the first revision - offer lower-clocked machines to get sales going at the lower-end and the 1.8 at the high end. Keep in mind that a 1.8 970 will be about at the performance level of a 3.6 G4.
Whoa, we have the same computer. Gotta love Ti 867s.
I personally hope I'll have the will power to resist the rev. A PPC 970. Rev. A's are notorious for having all sorts of flaws pop up in them (though not so much desktops as Powerbooks. That's one of the reasons your laptop is so sweet... rev. D, baby! No heat issues, no paint flaking issues (*knocks on wood*), no hinge-too-tight issues, etc. All the bugs are worked out...
The article also states the 1.8 - 2.5GHz range for the 970 AFTER it migrates to a 0.09µm process.
Can't seem to access the IBM press release any more, but I thought it mentioned the higher range still using a 0.13µm process.</strong><hr></blockquote>
Yeah, IBM pulled the press release... seems someone wasn't happy with that information getting out... nor the fairly hires screenshot of the blade server. Hint: if you still have it, try looking at it very closely.
Given the speed of the 970, or dual 970s at 2.5 gig and Altivec advantage...this product could be an adequate 'Virtual PC'.
And let's face it. It can only improve.</strong><hr></blockquote>
Okay, so now it likely emulates at about the speed of a 2 MHz Pentium IV.
Seriously, the entire paradigm that Bochs is using needs to be gutted for it to be effective, and something specifically coded for the PPC needs to be done in order for it to be even remotely usable (on par with VPC).
<strong>Yeah, IBM pulled the press release... seems someone wasn't happy with that information getting out... nor the fairly hires screenshot of the blade server. Hint: if you still have it, try looking at it very closely.</strong><hr></blockquote>
I nabbed the picture from the German IBM site. (Page 2 or 3 of this thread).
Hm, okay... Xilinx chips, one RAM slot (SD512-133), hard drive (IBM Thinkpad option ), a Lucent chip (???) under the hard drive, a couple of Broadcom chips, two PCI slots...
Wait a tick! That heat sink is awfully longish. Two CPUs? But I only see one "PIN #1"...
Okay, so now it likely emulates at about the speed of a 2 MHz Pentium IV.
Seriously, the entire paradigm that Bochs is using needs to be gutted for it to be effective, and something specifically coded for the PPC needs to be done in order for it to be even remotely usable (on par with VPC).</strong><hr></blockquote>
Don't worry, sooner or later, these guys that are recommending Bochs will eventually try to actually use it, and then they will shut up.
Oh wait, that's IBM opening up a fresh can of whup-ass! <hr></blockquote>
Moki. I looked at the pic. The heatsink looked very familiar... Was it that shape of the motherboard...? Or the fact it was using PC133 ram <img src="graemlins/lol.gif" border="0" alt="[Laughing]" />
As regards the emulation argument. Okay. But who would gut it and redefine it? Apple perhaps? If it's opensource and Unix...Apple have showed us what can be done
Emulation, smemulation. I think we may see Marklar before any built in emulation...
I've looked at the pic again. Now...all I'm seeing are dancing green dots... (Is the 'clue' subliminally revealed if I cross my eyes and stare for five minutes?)
It might just be me, but I think it looks a little like that travelstar is just lying on top of that board. Almost as if it was put there to cover something...
I am not an expert on Ram but isn't that rather slow DDRAM? Not the really fast stuff for its frontside bus. Maybe someone else could explain what this memory is.
Also I could not see the crystal for the processor chip(s?). Could it be under the heatsink too?
Comments
<strong>
I can't believe that Steve has that much influence over IBM. It is very surprising that Apple was able to get IBM to remove this information. Especially since Apple has not announced it is even using this chip. To me this is even more evidence that Apple will be using it. IBM doesn't have anywhere near the concern that Apple does for preannouncing products. Only Apple would care that much to have the information removed.</strong><hr></blockquote>
This strongly suggests that Apple has subsidised the development of this chip and has legally reserved certain rights about the publicity surrounding it. It may be just a case of the left hand not knowing what the right hand was doing - this was the IBM Germany site, not USA. They may not have been as on-the-ball as they might have been, and posted info that technically trampled on Apple's rights without realizing it. Or, it may be as simple as that they didn't get clearance from Apple before posting it - nothing inherently wrong with the info, they just didn't go through the right channels and get all the signatures they needed.
I'm beginning to suspect that the 970 was actually Apple's idea in the first place. They probably pitched it to Big Blue as one they could both use, and offered to fund much of its development. Curiouser and curiouser...
Yeah, there is so much optimism going around that if reality doesn't match up with all of our hopes, there is sure to be much dissapointment, and maybe anger in some (eg "I'm switching to Windows because the new machines are ONLY 1.8Ghz! That was supposed to be the LOW end!"). We're all being built up by press releases and 970 discussion in the forums. The higher you get, the harder you hit when you fall. I can't help but think about that alongside all of the excitement.
<strong>
I think hardcore gamers who aren't completely attached to Windows would switch for the performance, if the games were there. But I'm not so sure that they will be. Remember we still only have %3.
I don't think it's fair to say that the people who have moved to Windows moved simply because of performance. And I don't think its fair to say that everyone would suddenly switch when Apple's machines are the ones doing the ass kicking. Yes some people would, because they rely on high performance to get their work done to make the money to put the food on the table, and I'm not denying it. But plenty of people use computers for every day things and will be content with slower and cheaper PCs.
[ 03-01-2003: Message edited by: FrostyMMB ]</strong><hr></blockquote>
I think we're talking past each other a bit here. Let me try one more time, then I'll drop the subject.
I'm basing my opinion on two assumptions:
1) That the 970 offers superior enough performance that it can offer an emulated x86 environment as good as native x86 hardware - no significant performance penalty.
2) That there is a large number of Windows users who would be willing to switch to OS X were it not for their large investment in Windows software. This number would be enough to boost Apple's marketshare to the 10% range or higher.
My scenario then is that Apple offers a good IA-32 emulator within OS X. These potential switchers go to buy new hardware, and realise that they can get a Mac system that is as good as any x86 box to run their Windows software, and gets them OS X to boot. So they buy a Mac and load all their old Windows software onto it. Now, as the need arises for new versions of their software, what are they going to want? Another Windows version to run under emulation, or a native OS X version that will run significantly faster on their 970 system (even with crappy porting)? I don't think there's any question that developers will feel enormous pressure to get quality OS X native versions out. Those that don't respond will find their marketshare dropping as buyers head for their competitors.
I visualize the IA-32 emulation as being a lot like Classic mode. Crucially important to switchers for 6 months to a year, then gradually fading to almost complete disuse.
You don't have to agree with me, but that's the nuts and bolts of my argument. As I said at the start, I'll leave this topic alone now.
[quote]The hardware's purpose is simply to run the software without which the hardware has little utility other than as a decorative object or paper weight. <hr></blockquote>
Ask yourself this...
Can Windows run on anything other than some X86 processor-based system?
FrostyMMB writes:
[quote]So what? It's %3 of the market and they are still making the other %97 content. Quark is still around, but the Mac community is very angry with them. <hr></blockquote>
Um, how much do you want to bet that the majority of companies running Quark, are running it on a Mac and not some Windows crate. In certain markets, the PC has FAR LESS marketshare.
[quote]No, graphics (or video, or 3D) professionals most certainly would not. But the developers can save money and satisfy the large majority of Windows users. <hr></blockquote>
It isn't that simple... Besides, they should be diversifying.
[quote]The Mac marketshare is so small that it wouldn't make a huge impact on the success of the developer. Again, they are still making the large majority happy. <hr></blockquote>
Whoops! What it. You're making the same mistake again... You are mistakenly assuming that the Macs *overall* marketshare is am accurate representation of their marketshare in specific markets. Don't be silly. And don't forget, Apple is developing their X11 That too will aid in further development ;-) And then you are forgetting about all that continued chaos and uncertainty that will exist in the Windows world... It doesn't look like they will be getting 64-bit desktops for quite some time (end of decade?). However, what they *will* be getting is all that fancy DRM stuff . there seems to be an anti-Microsoft air going around lately. I think it's just gonna get worse.
On the marketshare thing you keep mentioning.. 3-4% could amount to market 40,000,000 potential users. That's nothing to ignore if your a developer. So, your 3% is rather misrepresented. If I were you, I'd think it through again ;-)
Still, I would find it rather disturbing knowing that my hard-earned $$ is being spent on apps coded by developers who are choosing the most ubiquitous (and arguably worst of the bunch) instead of a (possible) better option? And it's like saying that Goodyear will (should?) only make tires for the Toyota Carolla and Ford Escort because they hold the largest marketshare..
To quote my friend John C. Welch:
[quote]Neither is Maya, or Lightwave, or MatLab. But all of those either run, or are being ported to Mac OS X. The cost of the software is not an issue unless you are writing games. Remember, If it's a good tool, people pay for it. I don't think that Maya is going to grab 40 million users, but they'll get enough. MatLab is going to be an XFree86 application, ala IDL/Envi, so the port from the Unix versions of that application are pretty simple.
... I find it odd that an OS that is obviously on the way out, Irix, is considered a more viable option than Mac OS X for various reasons. - Jon C. Welch<hr></blockquote>
And Dave K. Every says:
[quote]We could just as readily ask then why does Microsoft do Office or Outlook for the Mac, etc... Because it is a very profitable segment for them. In fact, THOSE apps required complete rewrites and were much more major development efforts than what we are talking about here (since they used the Mac API's)... That's the job of software companies; create solutions to gain new markets. If you stop innovating, then you are stagnant and just waiting to be obviated by someone with more a clue (more vision) than you have.
When you port to other platforms, and gain new customers, they give you insights into new features or ways of thinking that help your entire market. This is one of the reasons why Microsoft keeps their Mac products; it makes their Windows versions BETTER. In fact, if I wasn't for the Mac products, I don't think most of the Windows apps would have been nearly as big. (Heck, most were Mac first). - (David K. Every) <hr></blockquote>
The Mac shouldn't be viewed as just "the other platform"
A significant point that is often overlooked is the fact that there is in fact a "crossplatform market". For instance, you cold look at it as three (or more) separate markets... The Mac market, the PC market and the crossplatform market (the Mac+PC market). That is the Mac+PC markets are markets where organizations use BOTH Mac *and* PC. So, If you support one platform (either one), you are actually missing TWO markets. 1. the other platform AND 2. the cross platform market. This indicates that ANY revenue generated from sales of supporting another platform (in this case Mac OS X/Unix) will also draw revenue from the *combined* market as well. How can you (or rather the developers) not see the potential benefits?
Historically, it's been *well* documented that it costs developers less to support the Mac than it does the PC. That means for every customer you convert to the Mac version you get a *complete upgrade* that you might not have sold (or at the very least a "competitive" upgrade, but the additional cost savings associated with supporting the Mac will provide additional revenue in the form of savings. Again, just another way to look at it. Most people and organizations only care to look at it from a single angle (narrow minded and shortsighted). Alias|Wavefront grew their Maya market 25% by deciding to support the Macintosh OS X platform. This should open some eyes. As a matter of fact it has, Mac development hasn't been better. and rumor is that DB/2 is the next major app headed in the direction of OS X.
[quote]Every company that is not paranoid about their markets, and thus stops trying to innovate (and penetrate new markets) in order to maximize profits of the now, sells their future for the present. This short-term quarterly report thinking is what has allowed not only companies but whole industries to be eaten alive.
Some of these companies do survive. They go into innovation through acquisition mode; and can sometimes acquire fast enough to tread water or break even. (Or at least slow the inevitable descent into oblivion). Most go under, or become pathetic shadows of what they once were, and shameful embarrassments compared to their potential.
So, these developers are choosing to *only* support Windows, but are laying the foundation for their own replacement. New platforms coming up are ignored, by them, but not their competition. They are entrenched in their current market and profit taking, and so are missing out on many opportunities that their competition isn't. Let's not learn the lessons of Novell, Lotus, DEC, Wang, DataGeneral, Packard Bell, and so on. Let's mimic them, because many of them were profitable, in the short term, right before they focused on the now (too long) and let the competition pass them by, and they disappeared. ... I tend to think that if they ignore the #2 OS, and the #1 UNIX OS, and the *fastest growing OS* (and UNIX OS), then they are not exactly being wise or planning for the future.
And you have to remember, ignoring that trend means more problems not only for Mac market, but also for all organizations that are mixed environments.
Remember, the tools are the platform. If they don't support my platform, then they are not supporting me (as I would like). This means that they will either drive customers to competing products (and as you know it is 10 times harder to get someone back once they've learned something else), or they will be telling me that my concerns don't matter to them (generating bad will). Neither is good business.
"Yet another platform" that is bigger than all but Windows (and has a few million seats). It seems like they should choose their platforms to reflect market trends, don't you think? - Dave K. Every <hr></blockquote>
Just some things to consider...
--
Ed M.
<strong>Only Apple would care that much to have the information removed.</strong><hr></blockquote>
Not quite. IBM would like to see it pulled if it was factually inaccurate. As in, if the quoted speed ranges are not in line with expected results or yields.
<strong>
Not quite. IBM would like to see it pulled if it was factually inaccurate. As in, if the quoted speed ranges are not in line with expected results or yields.</strong><hr></blockquote>
...and/or the prototype they showed had nothing to do with 970.
What if another white lie was power consumption? This is a 0.13 process, low-k dielectric, SOI, copper process, is it not? That's what East Fishkill (what a gross name) was all about: implementing ALL of IBM's process technologies to create chips that a) perform like crazy and b) draw less power.
0.9 is next, and coming fast from the sound of it. Man, 64-bit Powerbook 3D workstations are going to roXxor.
Just a thought.
edit: No one's going to complain (except Intel, hee hee) if IBM under-promises and over-delivers, right? No SEC investigations for stock manipulation or anything.
And the low-k dielectric & Silicon On Insulator (SOI) are a two way street: either faster speed chips drawing acceptablly low power, or hella fast chips drawing crazy power (SOI & LKD redue 'cross talk' AKA electrical interference by insulating everything, increasing reliability of signals).
either way,
*crrrrunnnk* **Krunnnnkkk** **cccrrrruuunnnkkk***
What's that grinding sound?
Oh wait, that's IBM opening up a fresh can of whup-ass!
[ 03-02-2003: Message edited by: 1337_5L4Xx0R ]</p>
Merry dreams!
[ 03-02-2003: Message edited by: fred_lj ]</p>
What are heatsinks made from again?
That's right... Aluminum!
So the Powerbook IS the heatsink...
Anyway, back to the 970. 1.8 GHz will be wonderful if we get it this fall. I'd go in for one alongside this TiBook. I don't think Apple'd push the 2.5GHz procs in the first revision - offer lower-clocked machines to get sales going at the lower-end and the 1.8 at the high end. Keep in mind that a 1.8 970 will be about at the performance level of a 3.6 G4.
I personally hope I'll have the will power to resist the rev. A PPC 970. Rev. A's are notorious for having all sorts of flaws pop up in them (though not so much desktops as Powerbooks. That's one of the reasons your laptop is so sweet... rev. D, baby! No heat issues, no paint flaking issues (*knocks on wood*), no hinge-too-tight issues, etc. All the bugs are worked out...
... I hope...
<strong>
The article also states the 1.8 - 2.5GHz range for the 970 AFTER it migrates to a 0.09µm process.
Can't seem to access the IBM press release any more, but I thought it mentioned the higher range still using a 0.13µm process.</strong><hr></blockquote>
Yeah, IBM pulled the press release... seems someone wasn't happy with that information getting out... nor the fairly hires screenshot of the blade server. Hint: if you still have it, try looking at it very closely.
<strong>200% increase is no slouchy.
Most basic software should be 'useable'.
Given the speed of the 970, or dual 970s at 2.5 gig and Altivec advantage...this product could be an adequate 'Virtual PC'.
And let's face it. It can only improve.</strong><hr></blockquote>
Okay, so now it likely emulates at about the speed of a 2 MHz Pentium IV.
Seriously, the entire paradigm that Bochs is using needs to be gutted for it to be effective, and something specifically coded for the PPC needs to be done in order for it to be even remotely usable (on par with VPC).
<strong>Yeah, IBM pulled the press release... seems someone wasn't happy with that information getting out... nor the fairly hires screenshot of the blade server. Hint: if you still have it, try looking at it very closely.</strong><hr></blockquote>
I nabbed the picture from the German IBM site. (Page 2 or 3 of this thread).
Hm, okay... Xilinx chips, one RAM slot (SD512-133), hard drive (IBM Thinkpad option
Wait a tick! That heat sink is awfully longish. Two CPUs? But I only see one "PIN #1"...
PCI slots on a blade server?
Screed
[ 03-02-2003: Message edited by: sCreeD ]</p>
<strong>
Okay, so now it likely emulates at about the speed of a 2 MHz Pentium IV.
Seriously, the entire paradigm that Bochs is using needs to be gutted for it to be effective, and something specifically coded for the PPC needs to be done in order for it to be even remotely usable (on par with VPC).</strong><hr></blockquote>
Don't worry, sooner or later, these guys that are recommending Bochs will eventually try to actually use it, and then they will shut up.
What's that grinding sound?
Oh wait, that's IBM opening up a fresh can of whup-ass! <hr></blockquote>
Moki. I looked at the pic. The heatsink looked very familiar... Was it that shape of the motherboard...? Or the fact it was using PC133 ram <img src="graemlins/lol.gif" border="0" alt="[Laughing]" />
As regards the emulation argument. Okay. But who would gut it and redefine it? Apple perhaps? If it's opensource and Unix...Apple have showed us what can be done
Emulation, smemulation. I think we may see Marklar before any built in emulation...
Lemon Bon Bon :cool:
Lemon Bon Bon
<strong>
Wait a tick! That heat sink is awfully longish. Two CPUs? But I only see one "PIN #1"...</strong><hr></blockquote>
If you look above the heatsink, you'll see a set of capacitors and VRMs...times two.
Also I could not see the crystal for the processor chip(s?). Could it be under the heatsink too?