I think it would be safe to assume this chip WILL be backwards compatible with 32bit instructions. PowerPC is a 64bit platform but many early chips (601-7455) have only used the 32bit extensions. Unless the processor was highly specialized for 64bit instructions only, a properly designed 64bit PowerPC will execute 32 and 64bit with no penalties. And memory addressing at 64bit would not affect it either.
Apple needs to stay with the current pricing strategy. They could probably put in a higher end system, but they cant get rid of the 1599 bottom end if they still want to double their market share. In fact they should be striving to get it back down to 1499, and possibly adding a "headless" consumer computer to thier line-up.
As for cost... this CPU is NOT gonna be G4 cheap... EVERYONE should get that idea outta their heads right this second. This CPU is based on the Power4 and if Apple is gonna use it the cost will be higher.
Wanna play with the pros your gonna have to pay for it.
TANSTAAFL (look it up if you need to) <img src="graemlins/lol.gif" border="0" alt="[Laughing]" />
Dave</strong><hr></blockquote>
Uh, er, um, IBM explicitly states for desktops, it is in the presentation category for PC's. Just a passing thought. I could very well be wrong, but it appears this cpu is for desktops and will not be vastly more expensive than Apple' s current high end.
Like I said, I could be wrong, we'll know a lot more in October.
<strong>ok the news.com article suggest this is Not a dual core. So atleast in its first implimentaion it will be a single core.
Moki/Programmer how hard is it to rewrite code for 64 bit assuming this chip is not backward compatible with 32 bit?</strong><hr></blockquote>
Most people don't realize that the original PowerPC spec included a full definition of the 64-bit ISA, and that it is a simple extension of the 32-bit ISA including a mode for running 32-bit code. There has already been a 64-bit PowerPC -- the 620, back in about '96.
I'm also pretty much convinced that between all the talk of VMX on IBM sites and the IBM co-owning the patents, the IBM SIMD implementation will be AltiVec compatible. The main question in my mind is whether they take the opportunity while going to 64-bit (and requiring a software update to use it) to introduce AltiVec-II. As I've said before, just introducing AltiVec-II by itself doesn't make a lot of sense, but as part of the 64-bit spec it does make a lot of sense. Any 32-bit code would continue to use AltiVec, but any code that is updated to be specific to the new 64-bit processor anyhow might as well be able to take advantage of an enhanced vector engine (i.e. longer vectors and more types).
<strong>This doesn't make any sense. Why would Apple just skip a number, like going from a g4 to a g6, without using g5. If there is no g5 its because they will change names. Think about it.
Comparatively speaking it would be like Intel skipping p5 and calling their next processor a p6. Common sense says "IT doesn't make any sense!"</strong><hr></blockquote>
I say Apple is going to use a new name for this Power4 derived PowerPC chip. I say they are going to call it the X chip. After all - it's going in the X-serve and the soon to be renamed PowerMac, the X-Mac.
Most people don't realize that the original PowerPC spec included a full definition of the 64-bit ISA, and that it is a simple extension of the 32-bit ISA including a mode for running 32-bit code. There has already been a 64-bit PowerPC -- the 620, back in about '96.
I'm also pretty much convinced that between all the talk of VMX on IBM sites and the IBM co-owning the patents, the IBM SIMD implementation will be AltiVec compatible. The main question in my mind is whether they take the opportunity while going to 64-bit (and requiring a software update to use it) to introduce AltiVec-II. As I've said before, just introducing AltiVec-II by itself doesn't make a lot of sense, but as part of the 64-bit spec it does make a lot of sense. Any 32-bit code would continue to use AltiVec, but any code that is updated to be specific to the new 64-bit processor anyhow might as well be able to take advantage of an enhanced vector engine (i.e. longer vectors and more types).</strong><hr></blockquote>
I thought that it was the other way around; PowerPC was designed with 64bit implementation in mind and the 32bit was a subset so to speak. This could all just be semantics though. I totally agree on your assessment of the IBM SIMD unit (VMX). It was in fact the Altivec code name before we knew it as Altivec, and Apple could simply rebrand it Velocity Engine II. The similarities are too striking to be considered. Think about it; with the G4 developers for Linux (which IBM is pushing hard) have had time to work with Altivec. It's being incorporated into the compilers. Why would IBM scrap the work that has already been done for altivec and come up with an entirely new SIMD unit that is not at least compatible with Altivec calls? Reason says IBM's VMX is compatible with Motorolas Altivec.
<strong>ok the news.com article suggest this is Not a dual core. So atleast in its first implimentaion it will be a single core.
Moki/Programmer how hard is it to rewrite code for 64 bit assuming this chip is not backward compatible with 32 bit?</strong><hr></blockquote>
You generally don't rewrite any code for the 32->64 bit transition. What happens a lot of the time is that the kernel is the only code that needs to be rewritten at all -- then it dolls out address space and handles such things automatically.
In terms of your application -- let's say you want to address more than 32 bits of address space at a time (which is the only real advantage to 64 bit computing, other than large integer calculations, and possibly moving more data over the bus at a time), then you just recompile your code. If you never assumed that your data types were 32 bits in size, you don't need to change your code at all.
I will caution people a bit -- just going from a 32 bit processor to a 64 bit processor doesn't necessarily gain anything in terms of performance. Additional address space, sure -- but you don't suddenly double the performance of the CPU by doubling the "bitness".
So moki you are suggesting that for developers the transition to a 64bit chip like this is no big deal? NeXt already ran on 64bit hardware so prospects look pretty good then. Although Digidesign i think will bybass the mach kernal with their Protools 6.0 release in order to access the hardware directly and make it interact with their TDM Dsp cards. So their transition and others like them might be more difficult.
Also I remember hearing from a developer and I think I remember who but I don't want to name names..but anyway that developer was suggesting to wait for around October for fast apple hardware. When I heard this I thought it was odd because obviously they are releases powermacs in august so october would be way too early for a release. But maybe he just meant that this chip would be revealed then.
Anyway thanks for all the info Moki (aka my source )
<strong>Another deciding factor is how hard is it going to be for developers to support a 64bit chip. Altough we do not know if this chip will have backward 32bit compatibility I assume it will not because it is based on the power 4. </strong><hr></blockquote>
Huh? I thought I'd seen that the POWER4 contained complete 32 bit compatability? Did I misremember this?
[quote]<strong>Also many of you seem to assume that this will go into a a comparable powermac line however i don't understand how they could base a chip on the Power 4 and have it be as cost effective as the G4. Perhaps these will only be available at first in a new line of even SuperPowerMacs that will be in a higher price bracket than the current ones.</strong><hr></blockquote>
Well, POWER4 has been around quite a while now, so costs should be dropping in general....but, if you strip out extra cores, the insane amount of cache, and a few other bits and pieces that you don't need (in this implementation), I would think it would get pretty cost effective.
You might be right, in it not being as cheap as G4, but then the question is....would Apple pay a couple hundred more for a proccessor that would give them a performance lead for the next 12-18 months? Especially if it meant that they could raise the prices the same amount?
<strong>I will caution people a bit -- just going from a 32 bit processor to a 64 bit processor doesn't necessarily gain anything in terms of performance. Additional address space, sure -- but you don't suddenly double the performance of the CPU by doubling the "bitness".</strong><hr></blockquote>
Indeed -- I expect that when this machine arrives we'll be able to compile the same program for both 32-bit and 64-bit and compare the performance of the two versions... and (all other things being equal) the 32-bit version will come out faster.
Actually, it is possible that a 64 bit platform will be faster - not at existing things, but at new things.
I've heard rumblings for some time now about 64 bit color. It would serve the same purpose as 24-bit audio: Even if you know you're mastering down to 16 bit @ 44.1kHz, you want the extra precision while you're manipulating the sounds. So, even if you know you'll be producing 32 (or, for Hollywood film color fidelity, 36) bit colors, wouldn't it be nice to work in a 64 bit color space in Photoshop? When cameras, monitors and 1 picoliter printers show up that can do better than 24 bit color (that cost less than tens of thousands of dollars), digital art should be able to muster all the warmth and fidelity of traditional analog.
Oh, and LightWave and its ilk will finally run at a decent clip.
<strong>So moki you are suggesting that for developers the transition to a 64bit chip like this is no big deal? NeXt already ran on 64bit hardware so prospects look pretty good then. Although Digidesign i think will bybass the mach kernal with their Protools 6.0 release in order to access the hardware directly and make it interact with their TDM Dsp cards. So their transition and others like them might be more difficult.
Also I remember hearing from a developer and I think I remember who but I don't want to name names..but anyway that developer was suggesting to wait for around October for fast apple hardware. When I heard this I thought it was odd because obviously they are releases powermacs in august so october would be way too early for a release. But maybe he just meant that this chip would be revealed then.
Anyway thanks for all the info Moki (aka my source )</strong><hr></blockquote>
Maybe. But a question/comment.....we're all assuming that there is a new tower coming this month....why?
There isn't really any hard evidence of that. It could just as easily be an extra 2 month wait 'til the update.
<strong>Oh, and LightWave and its ilk will finally run at a decent clip. </strong><hr></blockquote>
Oh, how I wish. I heared the ATI FireGL X1 (or something) will take over rendering from Maya, Lightwave & Co. Though I couldn't care less for Maya, the prospect of LW rendering-on-a-card makes me giggle like a 14yr old girl!
"Oh, how I wish. I heared the ATI FireGL X1 (or something) will take over rendering from Maya, Lightwave & Co. Though I couldn't care less for Maya, the prospect of LW rendering-on-a-card makes me giggle like a 14yr old girl!"
There are many opinions, but from what we know now about IBM's new processor, a scenario must pass two tests to make sense.
1) Apple will announce a G5 processor before October 15th. There is no way Steve Jobs would let this processor be examined in detail before he gets a chance to tell us all about it. Maybe even this early revelation came sooner than planned. There is no rule about when a new processor can be discussed on that Forum. A G5 could be in pilot production now, for all we know.
2) The scenario must account for the difference in system interface between a G4 and the new G5. Likely, these cannot share a common motherboard, or support the same speed RAM, or interface to other system chips in the same way. If this is true, the current G4 and the G5 would not mix well in a PowerMac lineup, with low end and high end models.
One scenario that passes the tests is where Apple introduces a new high end product. It might be in a 2U or 3U rackmount enclosure, but there could be both a server and a workstation version of it. In the future, the G5 surely will find its way into PowerMacs, maybe after IBM is supplying G4s with the same system interface. For now, the PowerMacs may get speed bumps and enhancements of the motherboard. Such a scenario also allows for a more gradual moving away from Motorola, and using new IBM chips.
<strong>Actually, it is possible that a 64 bit platform will be faster - not at existing things, but at new things.
I've heard rumblings for some time now about 64 bit color. It would serve the same purpose as 24-bit audio: Even if you know you're mastering down to 16 bit @ 44.1kHz, you want the extra precision while you're manipulating the sounds. So, even if you know you'll be producing 32 (or, for Hollywood film color fidelity, 36) bit colors, wouldn't it be nice to work in a 64 bit color space in Photoshop? When cameras, monitors and 1 picoliter printers show up that can do better than 24 bit color (that cost less than tens of thousands of dollars), digital art should be able to muster all the warmth and fidelity of traditional analog.
Oh, and LightWave and its ilk will finally run at a decent clip. </strong><hr></blockquote>
The items you mention would be better done on the AltiVec unit in 32-bit integer or floating point, and the AltiVec unit has better instructions for doing those things (i.e. integer mul-add, saturate, packing/unpacking, etc).
Software which is actually using 64-bit integers will perform better in 64-bit mode, but that's about it. The difference should be relatively small, however.
If IBM is making a chip for Apple that is based on the 64-bit Power4, but modified for a desktop, you can assume the cost will be brought down to a desktop processor price point and that it will run 32-bit code just fine since desktops are currently 32-bit. You can't just jump into 64-bit.
A 64-bit processor will have very few immediate benefits, but it will open doors for things to be developed on OS X that maybe wouldn't have been worth the trouble before. It will open doors more than anything else.
OS X on X86. Well, that would be foolish to switch as moki stated. It is an archaic architecture. *IF Apple was going to use Intel processors I would imagine it would be an Itanium or something of that sort. However, I don't see OS X moving away from PPC for general purposes unless the PPC ship hits an iceberg and Apple has to use Pentiums as a lifeboat. Sure Apple's got internal builds running on x86. You always need contingency plans.
It's possible that OS X on x86 will be used in some specialized areas and markets, even if the main product lines remain PPC. Maybe some special purpose x86 unix workstations are needed somewhere. Apple would be glad to supply them. It's a relatively portable OS and it's already at least partially running on x86 anyway. In theory, if the developer tools and frameworks were ported to x86, a Cocoa app could be simply recompiled to run on X86 with little to no effort involved. The places where trouble would appear would be when dealing with endians or assembly code.
It sounds like Apple has some good motherboards coming for us now and in the future. PowerPC might just start making it's comeback next week.
Once Apple switches to the new IBM chip next year, I'd imagine we'll have 1.8Ghz PowerBooks and iMacs available. Hopefully they will perform respectably to the wintel competition at the time.
The new workstation from Apple, using the new IBM CPUs, WILL have UMA3 chipsets on the main logic boards...!!!
Come on folks! We all remember the days of speculating on the various Apple chipsets, the mythical Unified Motherboard Architecture (aka UMA), and how it would revolutionize the PowerMac lineup...!
Comments
Moki/Programmer how hard is it to rewrite code for 64 bit assuming this chip is not backward compatible with 32 bit?
<strong>...Also, Apple can't raise Mac prices much unless they have truly kicks ass performance.
[ 08-09-2002: Message edited by: Stoo ]</strong><hr></blockquote>
Apple needs to stay with the current pricing strategy. They could probably put in a higher end system, but they cant get rid of the 1599 bottom end if they still want to double their market share. In fact they should be striving to get it back down to 1499, and possibly adding a "headless" consumer computer to thier line-up.
<strong>
Okay... lets look at this with some logic...
As for cost... this CPU is NOT gonna be G4 cheap... EVERYONE should get that idea outta their heads right this second. This CPU is based on the Power4 and if Apple is gonna use it the cost will be higher.
Wanna play with the pros your gonna have to pay for it.
TANSTAAFL (look it up if you need to) <img src="graemlins/lol.gif" border="0" alt="[Laughing]" />
Dave</strong><hr></blockquote>
Uh, er, um, IBM explicitly states for desktops, it is in the presentation category for PC's. Just a passing thought. I could very well be wrong, but it appears this cpu is for desktops and will not be vastly more expensive than Apple' s current high end.
Like I said, I could be wrong, we'll know a lot more in October.
<strong>ok the news.com article suggest this is Not a dual core. So atleast in its first implimentaion it will be a single core.
Moki/Programmer how hard is it to rewrite code for 64 bit assuming this chip is not backward compatible with 32 bit?</strong><hr></blockquote>
Most people don't realize that the original PowerPC spec included a full definition of the 64-bit ISA, and that it is a simple extension of the 32-bit ISA including a mode for running 32-bit code. There has already been a 64-bit PowerPC -- the 620, back in about '96.
I'm also pretty much convinced that between all the talk of VMX on IBM sites and the IBM co-owning the patents, the IBM SIMD implementation will be AltiVec compatible. The main question in my mind is whether they take the opportunity while going to 64-bit (and requiring a software update to use it) to introduce AltiVec-II. As I've said before, just introducing AltiVec-II by itself doesn't make a lot of sense, but as part of the 64-bit spec it does make a lot of sense. Any 32-bit code would continue to use AltiVec, but any code that is updated to be specific to the new 64-bit processor anyhow might as well be able to take advantage of an enhanced vector engine (i.e. longer vectors and more types).
<strong>This doesn't make any sense. Why would Apple just skip a number, like going from a g4 to a g6, without using g5. If there is no g5 its because they will change names. Think about it.
Comparatively speaking it would be like Intel skipping p5 and calling their next processor a p6. Common sense says "IT doesn't make any sense!"</strong><hr></blockquote>
I say Apple is going to use a new name for this Power4 derived PowerPC chip. I say they are going to call it the X chip. After all - it's going in the X-serve and the soon to be renamed PowerMac, the X-Mac.
<strong>
Most people don't realize that the original PowerPC spec included a full definition of the 64-bit ISA, and that it is a simple extension of the 32-bit ISA including a mode for running 32-bit code. There has already been a 64-bit PowerPC -- the 620, back in about '96.
I'm also pretty much convinced that between all the talk of VMX on IBM sites and the IBM co-owning the patents, the IBM SIMD implementation will be AltiVec compatible. The main question in my mind is whether they take the opportunity while going to 64-bit (and requiring a software update to use it) to introduce AltiVec-II. As I've said before, just introducing AltiVec-II by itself doesn't make a lot of sense, but as part of the 64-bit spec it does make a lot of sense. Any 32-bit code would continue to use AltiVec, but any code that is updated to be specific to the new 64-bit processor anyhow might as well be able to take advantage of an enhanced vector engine (i.e. longer vectors and more types).</strong><hr></blockquote>
I thought that it was the other way around; PowerPC was designed with 64bit implementation in mind and the 32bit was a subset so to speak. This could all just be semantics though.
<strong>ok the news.com article suggest this is Not a dual core. So atleast in its first implimentaion it will be a single core.
Moki/Programmer how hard is it to rewrite code for 64 bit assuming this chip is not backward compatible with 32 bit?</strong><hr></blockquote>
You generally don't rewrite any code for the 32->64 bit transition. What happens a lot of the time is that the kernel is the only code that needs to be rewritten at all -- then it dolls out address space and handles such things automatically.
In terms of your application -- let's say you want to address more than 32 bits of address space at a time (which is the only real advantage to 64 bit computing, other than large integer calculations, and possibly moving more data over the bus at a time), then you just recompile your code. If you never assumed that your data types were 32 bits in size, you don't need to change your code at all.
I will caution people a bit -- just going from a 32 bit processor to a 64 bit processor doesn't necessarily gain anything in terms of performance. Additional address space, sure -- but you don't suddenly double the performance of the CPU by doubling the "bitness".
Also I remember hearing from a developer and I think I remember who but I don't want to name names..but anyway that developer was suggesting to wait for around October for fast apple hardware. When I heard this I thought it was odd because obviously they are releases powermacs in august so october would be way too early for a release. But maybe he just meant that this chip would be revealed then.
Anyway thanks for all the info Moki (aka my source
<strong>Another deciding factor is how hard is it going to be for developers to support a 64bit chip. Altough we do not know if this chip will have backward 32bit compatibility I assume it will not because it is based on the power 4. </strong><hr></blockquote>
Huh? I thought I'd seen that the POWER4 contained complete 32 bit compatability? Did I misremember this?
[quote]<strong>Also many of you seem to assume that this will go into a a comparable powermac line however i don't understand how they could base a chip on the Power 4 and have it be as cost effective as the G4. Perhaps these will only be available at first in a new line of even SuperPowerMacs that will be in a higher price bracket than the current ones.</strong><hr></blockquote>
Well, POWER4 has been around quite a while now, so costs should be dropping in general....but, if you strip out extra cores, the insane amount of cache, and a few other bits and pieces that you don't need (in this implementation), I would think it would get pretty cost effective.
You might be right, in it not being as cheap as G4, but then the question is....would Apple pay a couple hundred more for a proccessor that would give them a performance lead for the next 12-18 months? Especially if it meant that they could raise the prices the same amount?
<strong>I will caution people a bit -- just going from a 32 bit processor to a 64 bit processor doesn't necessarily gain anything in terms of performance. Additional address space, sure -- but you don't suddenly double the performance of the CPU by doubling the "bitness".</strong><hr></blockquote>
Indeed -- I expect that when this machine arrives we'll be able to compile the same program for both 32-bit and 64-bit and compare the performance of the two versions... and (all other things being equal) the 32-bit version will come out faster.
I've heard rumblings for some time now about 64 bit color. It would serve the same purpose as 24-bit audio: Even if you know you're mastering down to 16 bit @ 44.1kHz, you want the extra precision while you're manipulating the sounds. So, even if you know you'll be producing 32 (or, for Hollywood film color fidelity, 36) bit colors, wouldn't it be nice to work in a 64 bit color space in Photoshop?
Oh, and LightWave and its ilk will finally run at a decent clip.
<strong>So moki you are suggesting that for developers the transition to a 64bit chip like this is no big deal? NeXt already ran on 64bit hardware so prospects look pretty good then. Although Digidesign i think will bybass the mach kernal with their Protools 6.0 release in order to access the hardware directly and make it interact with their TDM Dsp cards. So their transition and others like them might be more difficult.
Also I remember hearing from a developer and I think I remember who but I don't want to name names..but anyway that developer was suggesting to wait for around October for fast apple hardware. When I heard this I thought it was odd because obviously they are releases powermacs in august so october would be way too early for a release. But maybe he just meant that this chip would be revealed then.
Anyway thanks for all the info Moki (aka my source
Maybe. But a question/comment.....we're all assuming that there is a new tower coming this month....why?
There isn't really any hard evidence of that. It could just as easily be an extra 2 month wait 'til the update.
(Not that I think this is likely, but.....)
<strong>Oh, and LightWave and its ilk will finally run at a decent clip.
Oh, how I wish. I heared the ATI FireGL X1 (or something) will take over rendering from Maya, Lightwave & Co. Though I couldn't care less for Maya, the prospect of LW rendering-on-a-card makes me giggle like a 14yr old girl!
Heheheheheheheh...giggle...
<img src="graemlins/embarrassed.gif" border="0" alt="[Embarrassed]" />
Lemon Bon Bon
1) Apple will announce a G5 processor before October 15th. There is no way Steve Jobs would let this processor be examined in detail before he gets a chance to tell us all about it. Maybe even this early revelation came sooner than planned. There is no rule about when a new processor can be discussed on that Forum. A G5 could be in pilot production now, for all we know.
2) The scenario must account for the difference in system interface between a G4 and the new G5. Likely, these cannot share a common motherboard, or support the same speed RAM, or interface to other system chips in the same way. If this is true, the current G4 and the G5 would not mix well in a PowerMac lineup, with low end and high end models.
One scenario that passes the tests is where Apple introduces a new high end product. It might be in a 2U or 3U rackmount enclosure, but there could be both a server and a workstation version of it. In the future, the G5 surely will find its way into PowerMacs, maybe after IBM is supplying G4s with the same system interface. For now, the PowerMacs may get speed bumps and enhancements of the motherboard. Such a scenario also allows for a more gradual moving away from Motorola, and using new IBM chips.
<strong>Actually, it is possible that a 64 bit platform will be faster - not at existing things, but at new things.
I've heard rumblings for some time now about 64 bit color. It would serve the same purpose as 24-bit audio: Even if you know you're mastering down to 16 bit @ 44.1kHz, you want the extra precision while you're manipulating the sounds. So, even if you know you'll be producing 32 (or, for Hollywood film color fidelity, 36) bit colors, wouldn't it be nice to work in a 64 bit color space in Photoshop?
Oh, and LightWave and its ilk will finally run at a decent clip.
The items you mention would be better done on the AltiVec unit in 32-bit integer or floating point, and the AltiVec unit has better instructions for doing those things (i.e. integer mul-add, saturate, packing/unpacking, etc).
Software which is actually using 64-bit integers will perform better in 64-bit mode, but that's about it. The difference should be relatively small, however.
A 64-bit processor will have very few immediate benefits, but it will open doors for things to be developed on OS X that maybe wouldn't have been worth the trouble before. It will open doors more than anything else.
OS X on X86. Well, that would be foolish to switch as moki stated. It is an archaic architecture. *IF Apple was going to use Intel processors I would imagine it would be an Itanium or something of that sort. However, I don't see OS X moving away from PPC for general purposes unless the PPC ship hits an iceberg and Apple has to use Pentiums as a lifeboat. Sure Apple's got internal builds running on x86. You always need contingency plans.
It's possible that OS X on x86 will be used in some specialized areas and markets, even if the main product lines remain PPC. Maybe some special purpose x86 unix workstations are needed somewhere. Apple would be glad to supply them. It's a relatively portable OS and it's already at least partially running on x86 anyway. In theory, if the developer tools and frameworks were ported to x86, a Cocoa app could be simply recompiled to run on X86 with little to no effort involved. The places where trouble would appear would be when dealing with endians or assembly code.
It sounds like Apple has some good motherboards coming for us now and in the future. PowerPC might just start making it's comeback next week.
Once Apple switches to the new IBM chip next year, I'd imagine we'll have 1.8Ghz PowerBooks and iMacs available. Hopefully they will perform respectably to the wintel competition at the time.
The new workstation from Apple, using the new IBM CPUs, WILL have UMA3 chipsets on the main logic boards...!!!
Come on folks! We all remember the days of speculating on the various Apple chipsets, the mythical Unified Motherboard Architecture (aka UMA), and how it would revolutionize the PowerMac lineup...!
<img src="graemlins/smokin.gif" border="0" alt="[Chilling]" /> <img src="graemlins/smokin.gif" border="0" alt="[Chilling]" /> <img src="graemlins/smokin.gif" border="0" alt="[Chilling]" /> <img src="graemlins/smokin.gif" border="0" alt="[Chilling]" /> <img src="graemlins/smokin.gif" border="0" alt="[Chilling]" /> <img src="graemlins/smokin.gif" border="0" alt="[Chilling]" /> <img src="graemlins/smokin.gif" border="0" alt="[Chilling]" /> <img src="graemlins/smokin.gif" border="0" alt="[Chilling]" />
What I really want to see:
Dual IBM CPUs (single cores)
DDR333 SDRAM
nVidia/Apple graphics card (dual nv30 GPUs, massive amounts of DDRII RAM <512MB> & DUAL ADC ports)
Add any buzzwords or pipedreams as you see fit...!
Cheers!
Since people say you cannot double-pump the MPX bus, is it possible to have a seperate bus for each processor on a dual processor machine?