<strong>Adding 600Mhz and rolling out a new laptop architecture in one year is a lot more than either Moto or IBM have done in recent years.</strong><hr></blockquote>
Again, those 600 Mhz is only a 19.6% increase.</strong><hr></blockquote>
Who cares about percentage. The only reason the percentage is not as high as the G4 update is because the G4 clock rating is so low compared to the intel update. Who cares if I get a yugo with a 200% increase in HP if it still doesn't compete with a 10% increase the corvette? I'm with Matsu on this one!
Who cares about percentage. The only reason the percentage is not as high as the G4 update is because the G4 clock rating is so low compared to the intel update. Who cares if I get a yugo with a 200% increase in HP if it still doesn't compete with a 10% increase the corvette? I'm with Matsu on this one!</strong><hr></blockquote>
Would Intel be able to make an 600 MHz increase if they started at 1 GHz?
[[[Will Wintelon go IA64 or X86-64 ??? Intel will be ready in either case, you'll see.]]]
Itanic will not make it to the desktop in the next 5 years... It isn't meant for the desktop. Intel has said so. And where is Intel's X86-64 offering? Yamhill??
Intel seems to be in a serious predicament with respect to 64-bit... Then there are all the developers who would have to actually migrate their offerings and support YET ANOTHER version of Windows while maintaining legacy versions for at least another 5 to 6 years to appease today's customers and so on. Then there is compatibility issues, performance concerns, drivers... The nightmare seems to grow exponentially.
In all honesty, I think Intel and Microsoft are going to stick with 32-bit for the consumer line for some time to come... Possibly even the high-end creative-professional line as well. The 64-bit stuff will probably be aimed at servers. As a matter of fact, if Intel does this, it will limit where AMD will go with their 64-bit offerings as well. Who knows, AMD might end up being confined to the server market too if developers don't jump on developing their wares for a 64-bit version of Windows that runs on a 64-bit AMD chip which is completely different from a 64-bit Intel/Itanium chip.
This, in my mind is where the Wintelon platform seems to splinter....
Will M$ support two completely different CPU architectures and maintain compatibility across platforms? Intel is hell bent on making Itanium work... If that's the case and M$ chooses to develop for Intel, where does that leave AMD? Then there is the question of developer support...
Will developers be so quick as to code and support versions of their apps that run on 32-bit X86 as well as a 64-bit version? Now add in the possibility of having to code for two completely different architectures. This will only cause more confusion in the consumer sector.
People aren't upgrading now, what makes you think they will upgrade EVERYTHING *again* just to have 64-bit performance? Which processor will they choose?
The nice thing about PPC is that from the very beginning the PPC was designed to handle 64 and 32-bit. Again, just some things to think about...
The easy answer is, Yes. When 1-1.13Ghz P3's were out, the following twelve months saw the debut of 1.3-2.0Ghz P4's. Nearly double. The first examples (up to about 1.6Ghz) weren't even faster than the fastest P3 in many tasks but 2Ghz and up processors improved dramatically over their P3 predecessors, and the efficiency has improved with each re-working of the core. 2.4Ghz and up are very fast.
Would Intel be able to make an 600 MHz increase if they started at 1 GHz?</strong><hr></blockquote>
Don't understand your question. Don't work for Intel. What I do know is that the percentages are smaller for Intel's Mhz increases compared to Moto/IBM because their procs have a larger Mhz rating.
A 25% jump in the Intel Mhz proc....well lets not talk about that. That would just be crazy. I'll take two please.
No. It's not a 'great' chip. I don't think so, anyway. It's a triumph of dollars over design.
I think the AMD XP is a better chip. The Athlon was whoopassing the Pentium 4 for a large part of its life in fpu benches. I remember the first Athlon benches many moons ago slaughtering the Pentium opposition. And now? Well, AMD don't need the same Mhz to give the Pentium 4 a hard time. If the G4 was at 2 gig then I don't think the G4 would sweat it either. (But it's not at 2 gig-ish, is it? Hum.) I think the (don't quote me on this...) G4 is a better chip in some ways. Certainly more elegant but it's clearly ageing by some two years past its welcome.
But 'great' chip? I think the 970 is going to be a 'Great' chip. The 'Clawhammer' maybe too. But until we see it (970) in 'POWER'Macs, then who knows.
Matsu, DEFINATELY with you on this one. Intel always seem to come out of their 'walls' smelling of roses. When you're as dedicated to desktop chips as they are and have as much brute force and mullar dollar to throw at any given 'wall' they'll overcome the hurdles better than the like of...well Moto' for starters.
(G4 25% increase. Yeesh, please, let's not play the embarrassing loser's argument with Moto' involved in the same breath.)
Still, I've got a side bet with myself that the 970 is going to give any Intel chip next year a rough ride. So, ultimately, I don't really care.
Though I might if Steve Jobs gets on stage and gives us one of his periodic .Mac style 'pills'...
Well, I'll have to ask, what's so damn impressive about 600 MHz in a year, especially about a 20-stage pipelined ~3Ghz processor?
As told, it's only 19,6%.
When was the last time you did see a 19,6% MHz increase over a year? It's probably about as fast as when mot was trying to advance from 500MHz, with their G4's.
And, what makes you think IBM is unable to increase the MHz of the 970 by 600Mhz from when the (assumed) 1,8Ghz is released? One year, a 33% increase. I'd even bet you'd be pissed at IBM if that was all they managed, after all, it's less than mot's done with their G4's lately.
Dual800's, released summer/fall 2001 -> Dual 1250's, released summer/fall 2002 = ~1 year. Mhz increase = 450, which is 56%.
I'd say 450MHz is not that far from 600.
Percentage is relevant!
I just have to add I'd be very satisfied if IBM managed to even increase the MHz by 40% (->2,5Ghz), thrilled if they managed 56% as mot did (->2,8Ghz), but pissed if all was 33%, which is a lowly 600Mhz.
first of, bravo to Matsu for intelligently poking down the masses who want Intel to fail miserably.
Secondly, lets not forget that while the P4 may not be adding that much clock speed, the latest P4s will be far more efficient, meaning the mhz myth will be much less valid for the 3.6ghz P4s than it has been for the 1.x or 2.x ghz chips. apple/moto has been a disastor, and some people don't like to admit that for whatever reason.
If you want a faster computer, get a PC, there should be no argument to that
Would Intel be able to make an 600 MHz increase if they started at 1 GHz?</strong><hr></blockquote>
Perhapes not. but we have to ask why motorola is starting at 1 GHZ in the first place....
Motorola is slow. Real slow. their 25% increase may seems great in relative terms, but no one looks at relative terms. Everyone looks in term of absolute values and Motorola is definately not up there with Intel.
No. It's not a 'great' chip. I don't think so, anyway. It's a triumph of dollars over design.
I think the AMD XP is a better chip. The Athlon was whoopassing the Pentium 4 for a large part of its life in fpu benches. I remember the first Athlon benches many moons ago slaughtering the Pentium opposition. And now? Well, AMD don't need the same Mhz to give the Pentium 4 a hard time. If the G4 was at 2 gig then I don't think the G4 would sweat it either. (But it's not at 2 gig-ish, is it? Hum.) I think the (don't quote me on this...) G4 is a better chip in some ways. Certainly more elegant but it's clearly ageing by some two years past its welcome.
But 'great' chip? I think the 970 is going to be a 'Great' chip. The 'Clawhammer' maybe too. But until we see it (970) in 'POWER'Macs, then who knows.
The P4 3 Ghz has fine tech inside : net burst, HyperThreading. Hyperthreading give it a 20 % performance boost at equal mhz.
The Athlon XP is good, but he is now behind the P4 3 ghz in nearly all benchmarks. AMD need to release his BARTON or better his AThlon 64 soon to stay in the competition.
However i just ordered a Athlon XP 2400 : it's the best quality/prize ratio of the market.
However the PPC 970 will be a greater chip, which a much more room to grow in the future.
The G4 is not a match for a P4 3 ghz.
I don't see the need here , to say that Intel chips are crap : they have a design for mhz, but it works. They will soon reach limits, due to their high power consumption, but they are the faster design in the end of 2002.
The P4 3 Ghz has fine tech inside : net burst, HyperThreading. Hyperthreading give it a 20 % performance boost at equal mhz.
</strong><hr></blockquote>
How does HT work?
A multiprocessor system has two phyiscal chips. So each one can deal with data. Meaning proc A could run program A and proc B can run program B, both getting 100%. Or a multitheaded program could have threads running on both chips at 100%. A sinlge cored chip can only do the one task at a time, right? I mean, I am no engineer, but it seems like HT is what Intel is pushing, because it would be impossible for them to make a dual processor system (heat right?). So they came up with something (what is it, like super fast internal switching or something) that they can market as being faster at threaded apps.
Would a dual cored chip give better results than HT? Would a true dual proc system give better results than dual cored chip?
I just don't ge it. Could someone explain it to me? And while we are at it, what is netburst?
[edit] Isn't the 20% performance boost only for apps which are multithreaded? Or woudl it also improve performance of simultaneously running two apps (each would run a bit faster than running on a normal single proc system)?
look first of all if Intel doesn't tackle the heat issue the Mhz rating won't mean much. A P4 3.06 with HT enabled can consume 105 Watts. That's insane! How much efficiency could they possibly get from going to 90nm if they are also going to increase clock speed and keep HT?
Unless you yourself have seen the part in question or can provide a link that provides the power consumption figures for the 90nm P4's, you can't assume that they will be more efficient than the current chips.
Intel has never to my knowledge been able to put out a processor for portable use that is the same Proc running at the same speed as used in desktops. Apple on the other hand has been doing it for years. The rep of Mot might be taking a beating lately but there definitely has been an advantage for Apple using a chip supplier that also sells to the embedded hardware market.
As far as the progress that IBM has made in desktop processors, since when in recent memory has IBM had any reason to really push the performance of chips like the G3? That chip works fine in emedded hardware (including the Game Cube), and has made the iBook a uniquely powerful but small portable computer. There is a 1Ghz G3 available but Apple doesn't use it. Anybody want to venture a guess as to why?
The 970 is able to scale well by all indications and will be going to 90nm soon after production starts. I really think we have a winner in this chip.
A multiprocessor system has two phyiscal chips. So each one can deal with data. Meaning proc A could run program A and proc B can run program B, both getting 100%. Or a multitheaded program could have threads running on both chips at 100%. A sinlge cored chip can only do the one task at a time, right? I mean, I am no engineer, but it seems like HT is what Intel is pushing, because it would be impossible for them to make a dual processor system (heat right?). So they came up with something (what is it, like super fast internal switching or something) that they can market as being faster at threaded apps.
Would a dual cored chip give better results than HT? Would a true dual proc system give better results than dual cored chip?
I just don't ge it. Could someone explain it to me? And while we are at it, what is netburst?
[edit] Isn't the 20% performance boost only for apps which are multithreaded? Or woudl it also improve performance of simultaneously running two apps (each would run a bit faster than running on a normal single proc system)?
Intel has already have MP system with the Xeon. Xeon where the first HT chip selled by Intel. You can have both MP and HT working together.
Dual chip computer will work faster than HT obviously. But dual chip means twice the number of transistors, at the contrary to HT who only need 5 % more transistors. 20 % increase for a 5 % more transistors is a good deal.
Net burst is an internal feature of the P4 where the ALU work at twice the speed. A P4 3 ghz does Integer task at twice the speed = 6 ghz(for more info go to arstechnica).
HT rocks when the OS is running two soft together , for example typing under word and encoding a MP3 file. But many software are multhread in a way or an other. Soft like photoshop take advantages of this feature.
For dual cored chips, i think it's more performant than dual chip due to the high speed connection and lower latency between the two cores (the bandwitch and the latency are much higher between two separated chips).
Right that's why my car has a turbo-diesel. High torque low consomption. </strong><hr></blockquote>
Torque is not always better than HP. You can convert HP into torque by gearing, but never torque to HP. And torque is usually measured at the flywheel, not the wheels, where gearing is in effect.
But regardless, we all know the G4 sucks big time. And that the P4 is much better. So stop trying to make Intel look bad. They might be not so efficient and consume outrageos amounts of power, but in a desktop, there is nothing better right now.
Comments
<strong>Adding 600Mhz and rolling out a new laptop architecture in one year is a lot more than either Moto or IBM have done in recent years.</strong><hr></blockquote>
Again, those 600 Mhz is only a 19.6% increase.
<strong>
Again, those 600 Mhz is only a 19.6% increase.</strong><hr></blockquote>
Who cares about percentage. The only reason the percentage is not as high as the G4 update is because the G4 clock rating is so low compared to the intel update. Who cares if I get a yugo with a 200% increase in HP if it still doesn't compete with a 10% increase the corvette? I'm with Matsu on this one!
[ 12-12-2002: Message edited by: trailmaster308 ]</p>
<strong>
Who cares about percentage. The only reason the percentage is not as high as the G4 update is because the G4 clock rating is so low compared to the intel update. Who cares if I get a yugo with a 200% increase in HP if it still doesn't compete with a 10% increase the corvette? I'm with Matsu on this one!</strong><hr></blockquote>
Would Intel be able to make an 600 MHz increase if they started at 1 GHz?
Itanic will not make it to the desktop in the next 5 years... It isn't meant for the desktop. Intel has said so. And where is Intel's X86-64 offering? Yamhill??
<a href="http://www.arstechnica.com/archive/news/1033018263.html"; target="_blank">http://www.arstechnica.com/archive/news/1033018263.html</a>;
Intel seems to be in a serious predicament with respect to 64-bit... Then there are all the developers who would have to actually migrate their offerings and support YET ANOTHER version of Windows while maintaining legacy versions for at least another 5 to 6 years to appease today's customers and so on. Then there is compatibility issues, performance concerns, drivers... The nightmare seems to grow exponentially.
In all honesty, I think Intel and Microsoft are going to stick with 32-bit for the consumer line for some time to come... Possibly even the high-end creative-professional line as well. The 64-bit stuff will probably be aimed at servers. As a matter of fact, if Intel does this, it will limit where AMD will go with their 64-bit offerings as well. Who knows, AMD might end up being confined to the server market too if developers don't jump on developing their wares for a 64-bit version of Windows that runs on a 64-bit AMD chip which is completely different from a 64-bit Intel/Itanium chip.
This, in my mind is where the Wintelon platform seems to splinter....
Will M$ support two completely different CPU architectures and maintain compatibility across platforms? Intel is hell bent on making Itanium work... If that's the case and M$ chooses to develop for Intel, where does that leave AMD? Then there is the question of developer support...
Will developers be so quick as to code and support versions of their apps that run on 32-bit X86 as well as a 64-bit version? Now add in the possibility of having to code for two completely different architectures. This will only cause more confusion in the consumer sector.
People aren't upgrading now, what makes you think they will upgrade EVERYTHING *again* just to have 64-bit performance? Which processor will they choose?
The nice thing about PPC is that from the very beginning the PPC was designed to handle 64 and 32-bit. Again, just some things to think about...
--
Ed
The easy answer is, Yes. When 1-1.13Ghz P3's were out, the following twelve months saw the debut of 1.3-2.0Ghz P4's. Nearly double. The first examples (up to about 1.6Ghz) weren't even faster than the fastest P3 in many tasks but 2Ghz and up processors improved dramatically over their P3 predecessors, and the efficiency has improved with each re-working of the core. 2.4Ghz and up are very fast.
[ 12-12-2002: Message edited by: Matsu ]</p>
<strong>
Would Intel be able to make an 600 MHz increase if they started at 1 GHz?</strong><hr></blockquote>
Don't understand your question. Don't work for Intel. What I do know is that the percentages are smaller for Intel's Mhz increases compared to Moto/IBM because their procs have a larger Mhz rating.
A 25% jump in the Intel Mhz proc....well lets not talk about that. That would just be crazy. I'll take two please.
No. It's not a 'great' chip. I don't think so, anyway. It's a triumph of dollars over design.
I think the AMD XP is a better chip. The Athlon was whoopassing the Pentium 4 for a large part of its life in fpu benches. I remember the first Athlon benches many moons ago slaughtering the Pentium opposition. And now? Well, AMD don't need the same Mhz to give the Pentium 4 a hard time. If the G4 was at 2 gig then I don't think the G4 would sweat it either. (But it's not at 2 gig-ish, is it? Hum.) I think the (don't quote me on this...) G4 is a better chip in some ways. Certainly more elegant but it's clearly ageing by some two years past its welcome.
But 'great' chip? I think the 970 is going to be a 'Great' chip. The 'Clawhammer' maybe too. But until we see it (970) in 'POWER'Macs, then who knows.
Matsu, DEFINATELY with you on this one. Intel always seem to come out of their 'walls' smelling of roses. When you're as dedicated to desktop chips as they are and have as much brute force and mullar dollar to throw at any given 'wall' they'll overcome the hurdles better than the like of...well Moto' for starters.
(G4 25% increase. Yeesh, please, let's not play the embarrassing loser's argument with Moto' involved in the same breath.)
Still, I've got a side bet with myself that the 970 is going to give any Intel chip next year a rough ride. So, ultimately, I don't really care.
Though I might if Steve Jobs gets on stage and gives us one of his periodic .Mac style 'pills'...
Lemon Bon Bon
<img src="graemlins/lol.gif" border="0" alt="[Laughing]" />
.........<hr></blockquote>
Well, I'll have to ask, what's so damn impressive about 600 MHz in a year, especially about a 20-stage pipelined ~3Ghz processor?
As told, it's only 19,6%.
When was the last time you did see a 19,6% MHz increase over a year? It's probably about as fast as when mot was trying to advance from 500MHz, with their G4's.
And, what makes you think IBM is unable to increase the MHz of the 970 by 600Mhz from when the (assumed) 1,8Ghz is released? One year, a 33% increase. I'd even bet you'd be pissed at IBM if that was all they managed, after all, it's less than mot's done with their G4's lately.
Dual800's, released summer/fall 2001 -> Dual 1250's, released summer/fall 2002 = ~1 year. Mhz increase = 450, which is 56%.
I'd say 450MHz is not that far from 600.
Percentage is relevant!
I just have to add I'd be very satisfied if IBM managed to even increase the MHz by 40% (->2,5Ghz), thrilled if they managed 56% as mot did (->2,8Ghz), but pissed if all was 33%, which is a lowly 600Mhz.
But even 600Mhz is closing the performance-gap.
Secondly, lets not forget that while the P4 may not be adding that much clock speed, the latest P4s will be far more efficient, meaning the mhz myth will be much less valid for the 3.6ghz P4s than it has been for the 1.x or 2.x ghz chips. apple/moto has been a disastor, and some people don't like to admit that for whatever reason.
If you want a faster computer, get a PC, there should be no argument to that
<strong>
Would Intel be able to make an 600 MHz increase if they started at 1 GHz?</strong><hr></blockquote>
Perhapes not. but we have to ask why motorola is starting at 1 GHZ in the first place....
Motorola is slow. Real slow. their 25% increase may seems great in relative terms, but no one looks at relative terms. Everyone looks in term of absolute values and Motorola is definately not up there with Intel.
Sad, but true
<strong>"Anyway the p4 3 ghz is a great chip."
No. It's not a 'great' chip. I don't think so, anyway. It's a triumph of dollars over design.
I think the AMD XP is a better chip. The Athlon was whoopassing the Pentium 4 for a large part of its life in fpu benches. I remember the first Athlon benches many moons ago slaughtering the Pentium opposition. And now? Well, AMD don't need the same Mhz to give the Pentium 4 a hard time. If the G4 was at 2 gig then I don't think the G4 would sweat it either. (But it's not at 2 gig-ish, is it? Hum.) I think the (don't quote me on this...) G4 is a better chip in some ways. Certainly more elegant but it's clearly ageing by some two years past its welcome.
But 'great' chip? I think the 970 is going to be a 'Great' chip. The 'Clawhammer' maybe too. But until we see it (970) in 'POWER'Macs, then who knows.
<img src="graemlins/lol.gif" border="0" alt="[Laughing]" /> </strong><hr></blockquote>
The P4 3 Ghz has fine tech inside : net burst, HyperThreading. Hyperthreading give it a 20 % performance boost at equal mhz.
The Athlon XP is good, but he is now behind the P4 3 ghz in nearly all benchmarks. AMD need to release his BARTON or better his AThlon 64 soon to stay in the competition.
However i just ordered a Athlon XP 2400 : it's the best quality/prize ratio of the market.
However the PPC 970 will be a greater chip, which a much more room to grow in the future.
The G4 is not a match for a P4 3 ghz.
I don't see the need here , to say that Intel chips are crap : they have a design for mhz, but it works. They will soon reach limits, due to their high power consumption, but they are the faster design in the end of 2002.
<strong>
The P4 3 Ghz has fine tech inside : net burst, HyperThreading. Hyperthreading give it a 20 % performance boost at equal mhz.
</strong><hr></blockquote>
How does HT work?
A multiprocessor system has two phyiscal chips. So each one can deal with data. Meaning proc A could run program A and proc B can run program B, both getting 100%. Or a multitheaded program could have threads running on both chips at 100%. A sinlge cored chip can only do the one task at a time, right? I mean, I am no engineer, but it seems like HT is what Intel is pushing, because it would be impossible for them to make a dual processor system (heat right?). So they came up with something (what is it, like super fast internal switching or something) that they can market as being faster at threaded apps.
Would a dual cored chip give better results than HT? Would a true dual proc system give better results than dual cored chip?
I just don't ge it. Could someone explain it to me? And while we are at it, what is netburst?
[edit] Isn't the 20% performance boost only for apps which are multithreaded? Or woudl it also improve performance of simultaneously running two apps (each would run a bit faster than running on a normal single proc system)?
[ 12-13-2002: Message edited by: kupan787 ]</p>
<strong>
It is inevitable that... a ceiling will be reached... just as consumer automobiles have reached a ceiling for HP.</strong><hr></blockquote>
That's ludicrous. They most certainly have not.
Unless you yourself have seen the part in question or can provide a link that provides the power consumption figures for the 90nm P4's, you can't assume that they will be more efficient than the current chips.
Intel has never to my knowledge been able to put out a processor for portable use that is the same Proc running at the same speed as used in desktops. Apple on the other hand has been doing it for years. The rep of Mot might be taking a beating lately but there definitely has been an advantage for Apple using a chip supplier that also sells to the embedded hardware market.
As far as the progress that IBM has made in desktop processors, since when in recent memory has IBM had any reason to really push the performance of chips like the G3? That chip works fine in emedded hardware (including the Game Cube), and has made the iBook a uniquely powerful but small portable computer. There is a 1Ghz G3 available but Apple doesn't use it. Anybody want to venture a guess as to why?
The 970 is able to scale well by all indications and will be going to 90nm soon after production starts. I really think we have a winner in this chip.
[ 12-13-2002: Message edited by: nebcon65 ]</p>
<strong>How does HT work?</strong><hr></blockquote>
<a href="http://www.anandtech.com/cpu/showdoc.html?i=1746&p=4"; target="_blank">http://www.anandtech.com/cpu/showdoc.html?i=1746&p=4</a>;
and
<a href="http://www.anandtech.com/cpu/showdoc.html?i=1576&p=1"; target="_blank">http://www.anandtech.com/cpu/showdoc.html?i=1576&p=1</a>;
<strong>That's ludicrous. They most certainly have not.</strong><hr></blockquote>
Besides, the HP isn't important, the torque is.
<strong>
Besides, the HP isn't important, the torque is.</strong><hr></blockquote>
Right that's why my car has a turbo-diesel. High torque low consomption.
<strong>
How does HT work?
A multiprocessor system has two phyiscal chips. So each one can deal with data. Meaning proc A could run program A and proc B can run program B, both getting 100%. Or a multitheaded program could have threads running on both chips at 100%. A sinlge cored chip can only do the one task at a time, right? I mean, I am no engineer, but it seems like HT is what Intel is pushing, because it would be impossible for them to make a dual processor system (heat right?). So they came up with something (what is it, like super fast internal switching or something) that they can market as being faster at threaded apps.
Would a dual cored chip give better results than HT? Would a true dual proc system give better results than dual cored chip?
I just don't ge it. Could someone explain it to me? And while we are at it, what is netburst?
[edit] Isn't the 20% performance boost only for apps which are multithreaded? Or woudl it also improve performance of simultaneously running two apps (each would run a bit faster than running on a normal single proc system)?
[ 12-13-2002: Message edited by: kupan787 ]</strong><hr></blockquote>
Intel has already have MP system with the Xeon. Xeon where the first HT chip selled by Intel. You can have both MP and HT working together.
Dual chip computer will work faster than HT obviously. But dual chip means twice the number of transistors, at the contrary to HT who only need 5 % more transistors. 20 % increase for a 5 % more transistors is a good deal.
Net burst is an internal feature of the P4 where the ALU work at twice the speed. A P4 3 ghz does Integer task at twice the speed = 6 ghz(for more info go to arstechnica).
HT rocks when the OS is running two soft together , for example typing under word and encoding a MP3 file. But many software are multhread in a way or an other. Soft like photoshop take advantages of this feature.
For dual cored chips, i think it's more performant than dual chip due to the high speed connection and lower latency between the two cores (the bandwitch and the latency are much higher between two separated chips).
I think the 970 will generate plenty of that.
Hold the road like glue.
Lemon Bon Bon
<strong>
Right that's why my car has a turbo-diesel. High torque low consomption.
Torque is not always better than HP. You can convert HP into torque by gearing, but never torque to HP. And torque is usually measured at the flywheel, not the wheels, where gearing is in effect.
But regardless, we all know the G4 sucks big time. And that the P4 is much better. So stop trying to make Intel look bad. They might be not so efficient and consume outrageos amounts of power, but in a desktop, there is nothing better right now.
Regards,
Ben