Was the P4 an 'engineering failure'?

[zsdersw]

No, it was not a failure. It sold well, it worked (not always the fastest, but not always the slowest either), and it provided for some of the features of Conroe and its successors.

 

[Munky]

I voted yes. I actually bought one of the early P4's (lousy Dell...), and if it wasn't for the intel marketing at work, I would have went with AMD much earlier than I did. In fact, I don't remember any successfull new architecture that was clock for clock slower than the one it replaced.

 

[apoppin]

Originally posted by: zsdersw
No, it was not a failure. It sold well, it worked (not always the fastest, but not always the slowest either), and it provided for some of the features of Conroe and its successors.

Dictionary.com Unabridged
fail·ure /'fe?ly?r/ Pronunciation Key -
?noun
1. an act or instance of failing or proving unsuccessful; lack of success: His effort ended in failure. The campaign was a failure.

intel's "campaign" *failed*

they aimed for 10Gz ... they *got* 3.8Ghz

is that success? ... dumping 90% of your R&D into a product that only gets 38% of it's target?

intel *admitted* failure by dumping it ... like a pretty but unfaithful --and expensive - girlfriend

 

[zsdersw]

Originally posted by: apoppin
intel's "campaign" *failed*

they aimed for 10Gz ... they *got* 3.8Ghz

is that success? ... dumping 90% of your R&D into a product that only gets 38% of it's target?

intel *admitted* failure by dumping it ... like a pretty but unfaithful --and expensive - girlfriend

The P4 is not an engineering failure, whether they aimed for 1GHz, 10GHz, or 100GHz.

Intel shelved the low-IPC/high-clockspeed design philosophy that Netburst represented when it became clear that process technology hindered the architecture to the point of making it totally uncompetitive and that a suitable alternative existed in Dothan/Yonah that could be improved upon to give us Conroe.

 

[apoppin]

that's their PR



if "process technology hindered the architecture " it IS an 'engineering failure' !

 

[Brunnis]

Originally posted by: apoppin
that's their PR



if "process technology hindered the architecture " it IS an 'engineering failure' !

Exactly. As I wrote earlier in this thread, part of the engineering is to anticipate any obstacles that would keep the design from reaching the set goals. In light of this, the Netburst line of CPUs ultimately was a failure from an engineering standpoint.

And I see that some people in this thread still can't make a distiction between a product failure and an engineering failure. The P4 was far from a product failure. From an engineering standpoint the P4 failed to meet several of the goals that were set out and it did not fail to meet these goals by a small margin either. Although it wasn't a total failure, I'd still call any design that falls this short of reaching its goal a failure.

 

[zsdersw]

Originally posted by: apoppin
if "process technology hindered the architecture " it IS an 'engineering failure' !

Originally posted by: BrunnisExactly. As I wrote earlier in this thread, part of the engineering is to anticipate any obstacles that would keep the design from reaching the set goals. In light of this, the Netburst line of CPUs ultimately was a failure from an engineering standpoint.

And I see that some people in this thread still can't make a distiction between a product failure and an engineering failure. The P4 was far from a product failure. From an engineering standpoint the P4 failed to meet several of the goals that were set out and it did not fail to meet these goals by a small margin either. Although it wasn't a total failure, I'd still call any design that falls this short of reaching its goal a failure.

No, you're both wrong. When the goals are unrealistic (and especially when they're set by the marketing dept. and not the engineers) the architecture and design isn't the failure, the goal is the failure.

The goal failed, not the design and not the architecture.

And even if you don't buy all of that, the failure to reach 10GHz doesn't make Netburst a failure because that's not all there is to Netburst and its engineering specifics.

 

[myocardia]

Originally posted by: zsdersw
The P4 is not an engineering failure, whether they aimed for 1GHz, 10GHz, or 100GHz.

Intel shelved the low-IPC/high-clockspeed design philosophy that Netburst represented when it became clear that process technology hindered the architecture to the point of making it totally uncompetitive and that a suitable alternative existed in Dothan/Yonah that could be improved upon to give us Conroe.

You should stop arguing with the people who know what they're talking about, zsdersw, since the emboldened portion of your words above are in fact an admittal of failure. And yes, that's exactly why Netburst was shelved, and also why it was a failure, for the most part. The one shining light of the Netburst architecture were the Northwood C's; both before and after the Northwood C's, Intel had nothing that was competitive. For a company that was worth at least 10 times as much money (the day the A64 debuted) as it's competitor, that's pretty pathetic.

 

[coldpower27]

Originally posted by: Brunnis

Originally posted by: apoppin
that's their PR



if "process technology hindered the architecture " it IS an 'engineering failure' !

Exactly. As I wrote earlier in this thread, part of the engineering is to anticipate any obstacles that would keep the design from reaching the set goals. In light of this, the Netburst line of CPUs ultimately was a failure from an engineering standpoint.

And I see that some people in this thread still can't make a distiction between a product failure and an engineering failure. The P4 was far from a product failure. From an engineering standpoint the P4 failed to meet several of the goals that were set out and it did not fail to meet these goals by a small margin either. Although it wasn't a total failure, I'd still call any design that falls this short of reaching its goal a failure.

That's because everyone has a different meaning of what engineering failure means. IF everyone can agree to what they mean by saying engineering failure then we might reach a concensus.

The NetBurst architecture was basically engineered to be a clockspeed demon and it was successful in this regard, just not quite as long as what was projected or hoped for, which is a miscalculation on the longevity of the architecture. The concept for high speed Low IPC does work, when you can factor out the heat issue, the P4 doesn't have a problem reaching high clockspeeds as shown by what you can do when under extreme cooling.

It seems from the post in this thread different people draw the line differently as to where "engineering failure" lies, some people say it a non functioning product, some a product that doesn't last as long as intended.

 

[coldpower27]

Originally posted by: myocardia

Originally posted by: zsdersw
The P4 is not an engineering failure, whether they aimed for 1GHz, 10GHz, or 100GHz.

Intel shelved the low-IPC/high-clockspeed design philosophy that Netburst represented when it became clear that process technology hindered the architecture to the point of making it totally uncompetitive and that a suitable alternative existed in Dothan/Yonah that could be improved upon to give us Conroe.

You should stop arguing with the people who know what they're talking about, zsdersw, since the emboldened portion of your words above are in fact an admittal of failure. And yes, that's exactly why Netburst was shelved, and also why it was a failure, for the most part. The one shining light of the Netburst architecture were the Northwood C's; both before and after the Northwood C's, Intel had nothing that was competitive. For a company that was worth at least 10 times as much money (the day the A64 debuted) as it's competitor, that's pretty pathetic.

Not really, a company the size of Intel is powerful enough so that they can redefine the battlefield as they desire. It took a really long time before the focus could shift way from pure clockspeed to actual performance.

The Northwood B were competitive in performance, though they were more expensive, but that is due to the fact that Intel marketed towards clockspeed and was the better known brand. So the premium could be justified.

The Athlon 64's vs Pentium 4's was still competitive as Intel had the HyperThreading advantage which gave them a smoother system for normal usage, as well as a media encoding advantage. They had lower gaming performance, and higher power consumption.

The goals of a corporation is to generate cash, it doesn't matter if Intel has 10 times or 100 times more resources available at their disposal what Intel showed was that actual performance wasn't a necessary requirement to reaching that goal. It could do that through other means. Intel doesn't have to compete in the battlefield that you wish them to which is the overall performance crown, they competed with marketing instead, and that as history shown was quite successful. They only need to look like they have the performance crown, which was what high clock speed did with NetBurst.

 

[apoppin]

isn't the dictionary definition of failure enough?

the P4 failed to meet intel's goals and expectations
[who cares what the fanboys think is 'successful'? ... MOOT point]
--hence INTEL dumped it like a hot Prescott

cancelled it

stopped working on it

failed to fix the Prescott

started working on *something else* with "promise" ... unlike dead NetBurst

their best engineers could not meet intel's expectations
--hence it is an engineering failure

 

[coldpower27]

Originally posted by: apoppin
isn't the dictionary definition of failure enough?

the P4 failed to meet intel's goals
[who cares what the fanboys think is 'successful'?]s
hence INTEL dumped it like a hot Prescott

Pentium 4 failed to meet some of Intel's goals, so it's up for major debate. Pure dictionary definitions don't work for this world as there are many shades of grey and the world is not black and white.

You mean people that don't share your views right? :disgust: There is no further need to continue development for the Pentium 4 as the goals have currently shifted to performance per watt, and the Pentium 4 is simply not designed for that so when your goals change you shift your development resources to something which will accomplish those goals. Hence development of Core micro-architecture.

It's not realistic is it to expect to continue development on something that doesn't accomplish the new goals, but that's completely justifiable as the Pentium 4 was never designed with those goals in mind anyway.

Not continuing development of NetBurst is not any sign of failure anyway.

 

[apoppin]

Originally posted by: coldpower27

Originally posted by: apoppin
isn't the dictionary definition of failure enough?

the P4 failed to meet intel's goals
[who cares what the fanboys think is 'successful'?]s
hence INTEL dumped it like a hot Prescott

Pentium 4 failed to meet some of Intel's goals, so it's up for major debate. Pure dictionary definitions don't work for this world as there are many shades of grey and the world is not black and white.

You mean people that don't share your views right? :disgust: There is no further need to continue development for the Pentium 4 as the goals have currently shifted to performance per watt, and the Pentium 4 is simply not designed for that so when your goals change you shift your development resources to something which will accomplish those goals. Hence development of Core micro-architecture.

It's not realistic is it to expect to continue development on something that doesn't accomplish the new goals, but that's completely justifiable as the Pentium 4 was never designed with those goals in mind anyway.

Not continuing development of NetBurst is not any sign of failure anyway.

that simply isn't how it happened

you have a fantasy view of intel's P4 ...

intel's engineers came up against a brick wall ... they were STOPPED COLD in their tracks ... they could NOT *continue* with P4 ... their engineers *failed* to come up with a solution and they ran out of time ... AMD was kicking the stuffing out of intel's overpriced CPUs. ... and they had NOWHERE to go with it

there were no improvements to be made ... they were taking their thousand dollar server chips with 2MB L3 cache and losing to $500 FX CPUs ...

they ran out of headroom ... they created an fire-breathing monster with no more Mhz to push out of a FAILED NetBurst

intel had *no choice* ... but to DROP p4 in favor of something that was not an abject failure

P4 had *no future* they had to cut their losses and cast it off like a tired old whore

"no future" is a failure to all but the most hardened fanboy
... and intel did NOT *plan* the P4 as an "interim" solution ... they lost a lot of market share due to P4's failure to compete.

 

[myocardia]

Originally posted by: coldpower27
Not really, a company the size of Intel is powerful enough so that they can redefine the battlefield as they desire. It took a really long time before the focus could shift way from pure clockspeed to actual performance.

The Northwood B were competitive in performance, though they were more expensive, but that is due to the fact that Intel marketed towards clockspeed and was the better known brand. So the premium could be justified.

The Athlon 64's vs Pentium 4's was still competitive as Intel had the HyperThreading advantage which gave them a smoother system for normal usage, as well as a media encoding advantage. They had lower gaming performance, and higher power consumption.

The goals of a corporation is to generate cash, it doesn't matter if Intel has 10 times or 100 times more resources available at their disposal what Intel showed was that actual performance wasn't a necessary requirement to reaching that goal. It could do that through other means. Intel doesn't have to compete in the battlefield that you wish them to which is the overall performance crown, they competed with marketing instead, and that as history shown was quite successful. They only need to look like they have the performance crown, which was what high clock speed did with NetBurst.

I think you're failing to see the difference between an engineering failure, and a marketing failure, as this entire post seems to point out. The question asked in the OP had absolutely nothing to do with the P4 being a marketing failure; it wasn't. It was, however, an engineering failure, at least IMO, and for once, I seem not to be in the minority.

 

[Matt2]

I got my AXP TBred-B 1700+ @ 2.2ghz for $50 brand new from the egg.

No CPU will ever come close to that CPU as far as bang for the buck goes.

I'd take it over my Northy 2.4B any day.

 

[apoppin]

Celeron 300 had the biggest band-for-buck of modern CPUs

--imo, of course ... coppermine/t-bred/tualatin notwithstanding

 

[myocardia]

Originally posted by: apoppin
Celeron 300 had the biggest band-for-buck of modern CPUs

--imo, of course ... coppermine/t-bred/tualatin notwithstanding

I do hope you mean the Celeron 300A, because the Celeron 300 was one of the worst cpu's Intel ever made, as far as performance per $.

 

[apoppin]

yep ... Celeron300A ... the original - forgettable - Celeron 300 was cacheless

http://en.wikipedia.org/wiki/Celeron

it surpassed the PII400 when o/c'd ... i believe 450Mhz was just an "average" O/C ... a "given"

Intel was well aware of the poor reputation of the original Celeron and determined not to make the same mistake twice, with the result that the new Mendocino core Celeron was a good performer from the outset. Indeed, most industry analysts regarded the first Mendocino-based Celerons as too successful?performance was sufficiently high to not only compete strongly with rival parts, but also to attract buyers away from Intel's high-profit flagship, the Pentium II.

The key to the new Celeron's performance was cache. Where the old model had no secondary cache at all, the new part included 128 KiB of L2 cache as part of the chip itself. Otherwise, it was identical. With a total of 19.2 million transistors (including cache) on a single chip, the Mendocino Celeron was difficult and expensive to manufacture, but Intel managed a flawless execution of an ambitious project.

The first Mendocino-core Celeron was clocked at a then-modest 300 MHz but was almost twice as fast as the old cacheless Celeron at the same clockspeed. To distinguish it from the old model, Intel called it the 300A. Although the other Mendocino Celerons (the 333 MHz part, for example) did not have an A appended, some people call all Mendocino processors "Celeron-A" regardless of speed.

The Mendocino Celeron was the first mass-market CPU to utilise on-chip L2 cache. On-chip cache is difficult to manufacture; especially L2 as more of it is needed to attain an adequate level of performance. A benefit of on-die cache is that it can be made to run much faster than individual off-chip cache chips. Contrast this with the other common cache arrangements at that time. Most CPUs used mainboard mounted or slot mounted secondary L2 cache, which was very easy to manufacture, cheap, and simple to enlarge to any desired size. Typical cache sizes were 512 KiB to 1 MiB, typical speeds 66 to 100 MHz. The Pentium II had a pair of moderately high-speed L2 cache chips mounted on a special-purpose board alongside the processor itself. This was expensive and imposed practical cache-size limits, but allowed it to be clocked faster. Typical size was 512 KiB, always running at 1/2 of the processor speed. The new Mendocino Celeron had only 128 KiB of cache, but ran it at full clock speed.

Although the Mendocino Celeron cache was rather small, its high clock speed more than overcame that handicap, and the Mendocino Celeron was a success, particularly with the enthusiast market. Overclockers soon discovered that, given a high-end motherboard, the Celeron 300A could run reliably at 450 MHz. This was achieved by simply increasing the Front Side Bus (FSB) speed from the stock 66 MHz to the 100 MHz spec of the Pentium II. At this speed, the Mendocino Celeron rivaled the fastest x86 processors available.

of course i had one

@500Mhz [edit ... or was that a 450 Celeron at 500+Mhz? ... my 300A might have *only* been 450Mhz ... now that i think about it, i always got just 'average' good OCs]

. . . i.e. PIII 600 @ 800Mhz... Tualatin 1.2Ghz @ 1.6Ghz ... P4 2.80c @ 3.31Ghz ... and now P4 3.4EE @ 3.74Ghz

it's been a long run of intel CPUs ... even before the 300a ... and even this "intel fanboy" has to *admit* the P4 was an engineering *disaster*

ultimately, anyway --for intel ... it started slow with Williamette and could not compete with PIII ... finally matured into the NW-b and NW-c which were excellent OC-ers ... the P4's brief zenith ... it's peak ... then Prescott ... a total disaster - even compared with NW, initially --and then intel just *stopped* ... admitting they didn't have the time or the resources to Fix it.
:Q

i know i was taken aback ... it was a big decision for intel to basically admit failure


edited again

 

[myocardia]

I had a pair of 366's @ 550. Man, I thought I was something. That was the last (only) 2P system I owned, and the last dual "core" system I owned, until I finally bought an X2, not all that long ago. That BX chipset sure was something, wasn't it?

 

[zsdersw]

Originally posted by: myocardia
You should stop arguing with the people who know what they're talking about

Probably, but guess what? That's *not* those two, nor is it you.

since the emboldened portion of your words above are in fact an admittal of failure. And yes, that's exactly why Netburst was shelved, and also why it was a failure, for the most part. The one shining light of the Netburst architecture were the Northwood C's; both before and after the Northwood C's, Intel had nothing that was competitive. For a company that was worth at least 10 times as much money (the day the A64 debuted) as it's competitor, that's pretty pathetic.

As coldpower pointed out, there has been no definition of "failure". The limitations of process technology are simply among the basic truths under which all microprocessors must live and in the case of Netburst, failed the architecture. The architecture didn't fail the process technology.

(I wonder how Netburst would've done on the 45nm process??)

Marketing failed by playing and losing the expectations game with regards to how fast Netburst would become. Now, if you're going to use a marketing failure/mistake as a demonstration of "engineering failure", then I'm not the one with any explaining/clarifying to do.

 

[magreen]

Originally posted by: zsdersw
As coldpower pointed out, there has been no definition of "failure". The limitations of process technology are simply among the basic truths under which all microprocessors must live and in the case of Netburst, failed the architecture. The architecture didn't fail the process technology.

(I wonder how Netburst would've done on the 45nm process??)

Marketing failed by playing and losing the expectations game with regards to how fast Netburst would become. Now, if you're going to use a marketing failure/mistake as a demonstration of "engineering failure", then I'm not the one with any explaining/clarifying to do.

That's just silly. Say I come up with an amazing engineering design for an architecture that exploits the properties of transparent aluminum transistors that would triple current cpu performance ("Hello computer"). Then the thing fails to be actualized since nobody can make transparent aluminum, and I claim it was a great design, it was the process that failed, and marketing that failed for convincing everybody my thing would be triple the speed of current cpus. That would be ridiculous -- the design failed because it failed to take into account the reality, and the task the design was designed to accomplish was not accomplished.

Similarly with the Netburst architecture.

 

[coldpower27]

Originally posted by: myocardia

Originally posted by: coldpower27
Not really, a company the size of Intel is powerful enough so that they can redefine the battlefield as they desire. It took a really long time before the focus could shift way from pure clockspeed to actual performance.

The Northwood B were competitive in performance, though they were more expensive, but that is due to the fact that Intel marketed towards clockspeed and was the better known brand. So the premium could be justified.

The Athlon 64's vs Pentium 4's was still competitive as Intel had the HyperThreading advantage which gave them a smoother system for normal usage, as well as a media encoding advantage. They had lower gaming performance, and higher power consumption.

The goals of a corporation is to generate cash, it doesn't matter if Intel has 10 times or 100 times more resources available at their disposal what Intel showed was that actual performance wasn't a necessary requirement to reaching that goal. It could do that through other means. Intel doesn't have to compete in the battlefield that you wish them to which is the overall performance crown, they competed with marketing instead, and that as history shown was quite successful. They only need to look like they have the performance crown, which was what high clock speed did with NetBurst.

I think you're failing to see the difference between an engineering failure, and a marketing failure, as this entire post seems to point out. The question asked in the OP had absolutely nothing to do with the P4 being a marketing failure; it wasn't. It was, however, an engineering failure, at least IMO, and for once, I seem not to be in the minority.

I am not failing in anything, I just have a different definition then you on what is an engineering failure, and the topic creator has his own definition of what it means as well. Well the opinions in this thread seem to be varied, so your also not in the majority either.

 

[coldpower27]

Originally posted by: magreen

Originally posted by: zsdersw
As coldpower pointed out, there has been no definition of "failure". The limitations of process technology are simply among the basic truths under which all microprocessors must live and in the case of Netburst, failed the architecture. The architecture didn't fail the process technology.

(I wonder how Netburst would've done on the 45nm process??)

Marketing failed by playing and losing the expectations game with regards to how fast Netburst would become. Now, if you're going to use a marketing failure/mistake as a demonstration of "engineering failure", then I'm not the one with any explaining/clarifying to do.

That's just silly. Say I come up with an amazing engineering design for an architecture that exploits the properties of transparent aluminum transistors that would triple current cpu performance ("Hello computer"). Then the thing fails to be actualized since nobody can make transparent aluminum, and I claim it was a great design, it was the process that failed, and marketing that failed for convincing everybody my thing would be triple the speed of current cpus. That would be ridiculous -- the design failed because it failed to take into account the reality, and the task the design was designed to accomplish was not accomplished.

Similarly with the Netburst architecture.

Theres a difference between never getting it out the door, and there is not much of an excuse for creating something that can never be produced from the get go. NetBurst can and was developed and brought to market, so it is a viable way of doing things. It's difficult to foresee as far into the future as the 90nm node considering the Pentium 4 was introduced back in 2000.

 

[zsdersw]

Originally posted by: magreen
That's just silly. Say I come up with an amazing engineering design for an architecture that exploits the properties of transparent aluminum transistors that would triple current cpu performance ("Hello computer"). Then the thing fails to be actualized since nobody can make transparent aluminum, and I claim it was a great design, it was the process that failed, and marketing that failed for convincing everybody my thing would be triple the speed of current cpus. That would be ridiculous -- the design failed because it failed to take into account the reality, and the task the design was designed to accomplish was not accomplished.

Similarly with the Netburst architecture.

If the engineers of Netburst really claimed they could get to 10GHz, then it was a failure on the part of the engineers who saw potential that didn't exist within the confines of the process technology. Until/unless you can provide proof to that effect, the claim of 10GHz was a marketing mistake and a marketing failure.. not a Netburst engineering failure.

 

[coldpower27]

Originally posted by: apoppin

Originally posted by: coldpower27

Originally posted by: apoppin
isn't the dictionary definition of failure enough?

the P4 failed to meet intel's goals
[who cares what the fanboys think is 'successful'?]s
hence INTEL dumped it like a hot Prescott

Pentium 4 failed to meet some of Intel's goals, so it's up for major debate. Pure dictionary definitions don't work for this world as there are many shades of grey and the world is not black and white.

You mean people that don't share your views right? :disgust: There is no further need to continue development for the Pentium 4 as the goals have currently shifted to performance per watt, and the Pentium 4 is simply not designed for that so when your goals change you shift your development resources to something which will accomplish those goals. Hence development of Core micro-architecture.

It's not realistic is it to expect to continue development on something that doesn't accomplish the new goals, but that's completely justifiable as the Pentium 4 was never designed with those goals in mind anyway.

Not continuing development of NetBurst is not any sign of failure anyway.

that simply isn't how it happened

you have a fantasy view of intel's P4 ...

intel's engineers came up against a brick wall ... they were STOPPED COLD in their tracks ... they could NOT *continue* with P4 ... their engineers *failed* to come up with a solution and they ran out of time ... AMD was kicking the stuffing out of intel's overpriced CPUs. ... and they had NOWHERE to go with it

there were no improvements to be made ... they were taking their thousand dollar server chips with 2MB L3 cache and losing to $500 FX CPUs ...

they ran out of headroom ... they created an fire-breathing monster with no more Mhz to push out of a FAILED NetBurst

intel had *no choice* ... but to DROP p4 in favor of something that was not an abject failure

P4 had *no future* they had to cut their losses and cast it off like a tired old whore

"no future" is a failure to all but the most hardened fanboy
... and intel did NOT *plan* the P4 as an "interim" solution ... they lost a lot of market share due to P4's failure to compete.

I disagree I have quite a good grasp of what Intel wanted to do with the P4, and it was originally an interim solution as Intel wanted to replace NetBurst with Itanium based CPU's once they got cheap enough to become mass market, but unfortunately that never occured for Intel so the NetBurst line was around much longer then anticipated.

Italizing and Bolding things doesn't change the validity of any argument, not to mention the dirty analogies as well, but I guess since your only means of actually trying to compete is attempting to make someone angry, that isn't unexpected. :disgust:

Intel's engineers came to a brick wall at the 90nm node not because the architecture couldn't work at high frequency, but it was thermally limited.

AMD's K8 CPU weren't doing much kicking till you got to Athlon 64x2 Manchester/Toledo vs the Pentium D on the 90nm node Smithfield. The Pentium 4 was regardless competitive in certain tasks against the Athlon 64 due to HyperThreading.

NetBurst was never designed for actual performance, but to sell clockspeed and it was weaker then Athlon 64 in gaming so Intel used cache which gives a significant boost in games to help it out, and your stretching the prices as the Athlon FX was introduced at 713 USD with the Gallatin Pentium 4 EE at 999USD. Considering the clockspeed mantra and the fact of Intel higher popularity is at work here, the higher price can be justified.

As shown by the 65nm process the Pentium 4 could be thermally tamed, but there was a lot of negative publicity from the 90nm node Pentium 4's so it was time to market based on something else entirely and hence now we have performance per watt. Basically it was time for Intel to redefine the battlefield for something new and fresh.

Athlon 64's impact took a very long time to be materialized, ever since the Athlon 64 was introduced to the point of Q4 2006 AMD gained 9.5% marketshare. Intel lost about 8% marketshare, with the difference being made up from VIA marketshare. So roughly a 10% loss, with the bulk of those gains happening after the Smithfield from Q3 2005 to Q4 2006 6.5% gain during that timeframe.