What Processor Brand will you choose as your next upgrade

[VASSA]

INTEL is the best

 

[Anusha]

well that almost sounds like a insult...it seems u never lived the stone age lool cuz one have to know the stone age to comment on present, AMD kicked intel ass for almost 5 years before intel came up with Core 2 (with the original Athlon then Athlon XP and then Athlon 64), remember if it wasnt for AMD we will still be cooking on Pentium 4's hell maybe less, AMD made intel to make core 2, they were almost always ahead of intel technologically (they made the first true dual core CPU, they were the first to include the memory controller on CPU die, they dropped old FSB and has been using HT for 5 years now, even tho the Phenoms are slower, technologically they are more advanced than Core 2 processors, and all intel do is babble abt moors law and do die shrinks every few months. so have some respect mate.

I was kinda thinking about this and figured this actually shows a weakness of AMD.

Look at it like this.

In the K8 days, when Intel were moving so fast from 90nm to 65nm without staying long enough in one process technology, I used to blame Intel, saying that they will run out of ability to miniaturize the chips after some point, as rapidly as they would want. AMD were optimizing their 90nm tech for a long time, and actually broke the 3GHz mark on the K8. Most people didn't think they will reach that far, especially on the dual core. But AMD really did drain the juice out of their 90nm tech. That was really admirable. (they released the 6400 X2 at 3.2GHz for crying out loud!)

Intel moves from one process tech to another fast because they can, but doesn't optimize one tech as much as AMD, as long as AMD. That is what I feel at least. However, we can't forget that Intel is still using lots of old technologies, such as FSB and MCH. Even without using the superior techs like HTT and IMC, Intel have managed to outperform a CPU that uses all those superior techs....and by a heavy margin as well.

Don't you think that is more admirable? Isn't that optimizing? Isn't that what should be done? I mean, bringing all these new technologies into the salad and if they can't really deliver the expected gains, why bother at all.

Athlon64 was a really good core. Why did they have to design Phenom if it is was slower clock to clock to the A64? They could have just moved to 65nm and release the A64 at 3.4, 3.6 and even higher clocks.

It is like paying for a Ferrari 430 compared to resurrecting a crappy Supra from the junk yard and optimizing it to perform better than the Ferrari. Though the Ferrari will sound great on paper (specs), it will not be any better than the Supra.

Just my thoughts......

 

[sameera_ranga]

INTEL is best one

 

[zCexVe]

I was kinda thinking about this and figured this actually shows a weakness of AMD.

Look at it like this.

In the K8 days, when Intel were moving so fast from 90nm to 65nm without staying long enough in one process technology, I used to blame Intel, saying that they will run out of ability to miniaturize the chips after some point, as rapidly as they would want. AMD were optimizing their 90nm tech for a long time, and actually broke the 3GHz mark on the K8. Most people didn't think they will reach that far, especially on the dual core. But AMD really did drain the juice out of their 90nm tech. That was really admirable. (they released the 6400 X2 at 3.2GHz for crying out loud!)

Intel moves from one process tech to another fast because they can, but doesn't optimize one tech as much as AMD, as long as AMD. That is what I feel at least. However, we can't forget that Intel is still using lots of old technologies, such as FSB and MCH. Even without using the superior techs like HTT and IMC, Intel have managed to outperform a CPU that uses all those superior techs....and by a heavy margin as well.

Don't you think that is more admirable? Isn't that optimizing? Isn't that what should be done? I mean, bringing all these new technologies into the salad and if they can't really deliver the expected gains, why bother at all.

Athlon64 was a really good core. Why did they have to design Phenom if it is was slower clock to clock to the A64? They could have just moved to 65nm and release the A64 at 3.4, 3.6 and even higher clocks.

It is like paying for a Ferrari 430 compared to resurrecting a crappy Supra from the junk yard and optimizing it to perform better than the Ferrari. Though the Ferrari will sound great on paper (specs), it will not be any better than the Supra.

Just my thoughts......

Well Anusha ppl have different thoughts.This is my point of view.I have somw different way of looking into things,pardon on that but just read it.
When AMD beat Intel hell out with the Athlon 64 core using LDT(HT),and IMC Intel wanted to come on top coz frankly they were going sales down rapidly.They needed a huge improvement over the core coz they never thought of that much competition from a company like AMD.Reaching higher clocks with what they had was impossible.But you cant give up coz you are intel,you need to keep it up.I have money,hell load but I cant have technology.SO quickest solution is to move to lesser nm tech,just to get the clocks higher and keep it up.And that time will be used for new tech.So then came the Core 2 line up from that time.But still the MCH+FSB.Then lucky for them AMD just went with new inventions considering far long term gains,which they will have in their predicted road map up until 2011.HT3.0,AM3 etc.But Intel got a lot time from core2 conroe,but still they were not finished with quick path(sad to say still 10GB/S slower at a link than HT)and IMC like AMD.Also the native quad L3 cache and things.Things were going again bad for Intel although sales were good,So came into 45nm and planned for 32nm-You pay us a lot -now they got more time and will release AIO package with nehalem hopefully will come by par with AMD technologies.Still AMD gains juice from 65nm,If they move into full time 45nm and as predicted,more into 32nm directly,AMD can beat Intel as they have technology at top notch.But still sadly AMD thought of intergrating GPU into CPU a whole new side and currently have research on it.This will look more brighter for them as I'm sure it will not be on X86,which they still pay Intel for that.Just my 0.02.
Edit:I just realized how sad it is to think about a CPU IRL ,more than the engineers at Intel or AMD.

 

[krizta]

INTEL

 

[Anusha]

Well Anusha ppl have different thoughts.This is my point of view.I have somw different way of looking into things,pardon on that but just read it.
When AMD beat Intel hell out with the Athlon 64 core using LDT(HT),and IMC Intel wanted to come on top coz frankly they were going sales down rapidly.They needed a huge improvement over the core coz they never thought of that much competition from a company like AMD.Reaching higher clocks with what they had was impossible.But you cant give up coz you are intel,you need to keep it up.I have money,hell load but I cant have technology.SO quickest solution is to move to lesser nm tech,just to get the clocks higher and keep it up.And that time will be used for new tech.So then came the Core 2 line up from that time.But still the MCH+FSB.Then lucky for them AMD just went with new inventions considering far long term gains,which they will have in their predicted road map up until 2011.HT3.0,AM3 etc.But Intel got a lot time from core2 conroe,but still they were not finished with quick path(sad to say still 10GB/S slower at a link than HT)and IMC like AMD.Also the native quad L3 cache and things.Things were going again bad for Intel although sales were good,So came into 45nm and planned for 32nm-You pay us a lot -now they got more time and will release AIO package with nehalem hopefully will come by par with AMD technologies.Still AMD gains juice from 65nm,If they move into full time 45nm and as predicted,more into 32nm directly,AMD can beat Intel as they have technology at top notch.But still sadly AMD thought of intergrating GPU into CPU a whole new side and currently have research on it.This will look more brighter for them as I'm sure it will not be on X86,which they still pay Intel for that.Just my 0.02.
Edit:I just realized how sad it is to think about a CPU IRL ,more than the engineers at Intel or AMD.

Well, the problem with your argument or AMD's vision is that, if their technologies will only give the full yields at the end of 2011, which you say is the end of K10 architecture, then the problem still remains, rather worsens. Intel will release their Nehalem at the end of this year and if they get it right, which I'm sure they will as they have prototypes running at 3.2GHz, then AMD will be again penalized for not using their full potential for the benefit of their products.

And one thing to note is that Intel's C2D didn't just appear out of nowhere. The Pentium M has been there since about 2 years prior to the C2D, and they could have brought it to the desktop at that time. They didn't, maybe because it required a separate socket and that is a big gamble to take as the 775 was just released. Anyway, that's all in history...

 

[fallenzeraphine]

well, u guys have some points, yes intel's Core 2 is faster even on a old Technology such as FSB simply because core2s have superior performing architecture than a K8, and simply none of these processors can utilize the full bandwidth of a HT bus completely. but my point is that AMD took that step, they pushed the envelope, i admire that quality of AMD, but as i see it AMD had one weakness after they created the highly successfull K8, they kinda forgot intel will strike back, the work on their new processor archtecture was slow, and when intel relased core 2, everything got really dark for AMD, they were just out resourced by intel, K10 is just a extention to K8 its not a new architechture at all, and AMD couldnt move to new fabrication processes cuz simply they dont have the resources to do so, but i like the spirit of AMD they were allways under pressure frm intel, and they always managed to bring out some quality product to push the envelope, i think AMD always made intel to create new processor archtechtures all along, even now AMD has really promising plans for the future like the Fusion processor, but its sad to see them go down, they are having a hard time right now, the Phenom was late and it cant go neck to neck with core 2, and they never had enough time or money to create a new architechture, im hoping the 45nm K10.1 parts will shed some light on AMD, untill they relase the all new Bulldozer architechture.

 

[Anusha]

well, u guys have some points, yes intel's Core 2 is faster even on a old Technology such as FSB simply because core2s have superior performing architecture than a K8, and simply none of these processors can utilize the full bandwidth of a HT bus completely. but my point is that AMD took that step, they pushed the envelope, i admire that quality of AMD, but as i see it AMD had one weakness after they created the highly successfull K8, they kinda forgot intel will strike back, the work on their new processor archtecture was slow, and when intel relased core 2, everything got really dark for AMD, they were just out resourced by intel, K10 is just a extention to K8 its not a new architechture at all, and AMD couldnt move to new fabrication processes cuz simply they dont have the resources to do so, but i like the spirit of AMD they were allways under pressure frm intel, and they always managed to bring out some quality product to push the envelope, i think AMD always made intel to create new processor archtechtures all along, even now AMD has really promising plans for the future like the Fusion processor, but its sad to see them go down, they are having a hard time right now, the Phenom was late and it cant go neck to neck with core 2, and they never had enough time or money to create a new architechture, im hoping the 45nm K10.1 parts will shed some light on AMD, untill they relase the all new Bulldozer architechture.

My question is, why is AMD investing more time and money on things that don't give out any performance benefits? If the current CPUs can't take the full advantage of the hypertransport bus, then AMD's efforts are invain. One might argue that they did it thinking about the future. But what the hell? How can afford to think about a future, when they already don't have a future unless they get the things that matter, right. What I understand is that they are wasting their time and resources on things that don't matter at all. They have to reassess their priorities.

And one thing I like about Intel is that they don't boast about things that they haven't yet accomplished. But look at AMD. They were so sure that their design was so superior to Intel Core 2 Duo and about 6 months before the K10 launch they were showing "hypothitical" benchmark scores to show how fast their K10 CPUs were. They even showed 3GHz overclock results, but up to now, "3GHz" is only for people who are "lucky" with overclocks.

On the other hand, we hardly knew about the Core 2 Duo release until about one month to go before launch, and even then they didn't go big on it. They let the reviewers speak for themselves. (Sure, they had Core Duo released few months back and that was a pretty good CPU. We could predict the final results of the Core 2 Duo as well.)

Anyways, looking at Intel Core 2 Duo, even at 65nm, they really got great yields out of the E6000 processors. zCexcVe might say that Intel used 45nm to boost their clock speeds, but what I think is their primary objective was to lower the power consumption. I mean, look at all the E6000 CPUs overclocking beyond 3.6GHz with ease. I believe they could have released 3.2GHz or even 3.4GHz CPUs officially on 65nm. But then the public would have moaned saying that the power consumption is too high. That's life. But the 65nm G0 stepping has really lowered the power consumption and they could have released a 3.2GHz or 3.4GHz CPU anyway....at least a dual core one.

 

[ufirst]

Intel

 

[niro123]

Intel

 

[zCexVe]

Anyways, looking at Intel Core 2 Duo, even at 65nm, they really got great yields out of the E6000 processors. zCexcVe might say that Intel used 45nm to boost their clock speeds, but what I think is their primary objective was to lower the power consumption. I mean, look at all the E6000 CPUs overclocking beyond 3.6GHz with ease. I believe they could have released 3.2GHz or even 3.4GHz CPUs officially on 65nm. But then the public would have moaned saying that the power consumption is too high. That's life. But the 65nm G0 stepping has really lowered the power consumption and they could have released a 3.2GHz or 3.4GHz CPU anyway....at least a dual core one.

My nick is with 6 letters I'm not saying that they aim only that.Big companies like Intel,nVIDIA,AMD never put their money just to have one gain.They have multiple primitive objectives.As saying on 45nm tech ,power consumption,higher clocks,less cost might have been their main interests.

 

[Billl Gates]

 

[Billl Gates]

 

[Shiwantha]

Intel

 

[GihanFX]

Intel forever

 

[Anusha]

My nick is with 6 letters I'm not saying that they aim only that.Big companies like Intel,nVIDIA,AMD never put their money just to have one gain.They have multiple primitive objectives.As saying on 45nm tech ,power consumption,higher clocks,less cost might have been their main interests.

That was just a typing mistake

 

[farshad]

no idea