Was the P4 an 'engineering failure'?

[magreen]

Originally posted by: 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.

But since the P4 was slower clock-for-clock than the P3 it replaced, how could the engineers have had any designs for P4's success other than ultra-high clock speeds?

 

[tno]

Magreen,

Thanks! I just have trouble believing that any company would choose to make the spectacularly huge investment necessary to create two chipset teams knowing that one would be making low yield mobile products whose toes you would step on by releasing P4 mobile products. Sure, maybe Intel told the Israeli team that if they couldn't prove this plan would work with some sort of product (the Pentium M) then they would be quietly shut down and all resources be devoted to either a NetBurst solution or a true P4 succesor.

As far as calls that the P4's are failures because they never met Intel's goal of 10Ghz, I think its safe to say that the architecture, if on well cooled and on a small enough architecture, could easily meet this goal. People were hitting close to 4Ghz on the .13 micron process so imagine what they could do on a process less than half that. But with their Ace in the Hole ready to go why wait.

Glad to be here,

Tno

 

[zsdersw]

Originally posted by: magreen
But since the P4 was slower clock-for-clock than the P3 it replaced, how could the engineers have had any designs for P4's success other than ultra-high clock speeds?

Of course they knew that clock speed was the key to Netburst's performance, but an arbitrary figure like 10GHz I'm fairly certain was something no engineer expected.. or claimed would be do-able.

 

[apoppin]

Originally posted by: 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.

after all *that* long-winded spiel, you really agree the P4 IS an engineering failure

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.

:roll:

some much for the "validity" of your argument ... self-contradictions are usually a clue ... "thermally limited" with NO solution = an 'engineering failure'

in ANY language and with an agreed-upon universal *dictionary* definition of FAILURE
:thumbsdown:

spin it all you want

you have no grasp of it at all ... i don't want to even bother to address the rest of your fantasy ... the main point is that the intel Engineers could not *fix* the P4's many problems including your self-admitted thermal problems ... so it was ABANDONED as a 'failure'.

 

[zsdersw]

Originally posted by: apoppin
you have no grasp of it at all ...

.. says the pot to the kettle.

 

[apoppin]

Originally posted by: zsdersw

Originally posted by: apoppin
you have no grasp of it at all ...

.. says the pot to the kettle.

at least the 'pot' can use Mr Kettle's own words to show he also agrees that the P4 is an 'engineering failure' ... he just refuses to acknowledge his own words

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.

intel's best engineers failed to *solve* the thermal limit "problem" of P4 and the project was canned because P4 was not competitive with FX-series - for a long time ... intel knew better then to pour money into an engineering failure as there was no *hope* for it ... barring a 'miracle'

how basic is that?
:roll:

unless you are in denial?

 

[zsdersw]

Originally posted by: apoppin
at least the 'pot' can use Mr Kettle's own words

No, you can't.

to show he also agrees that the P4 is an 'engineering failure' ... he just refuses to acknowledge his own words

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.

intel's best engineers failed to *solve* the thermal limit "problem" of P4 and the project was canned because P4 was not competitive with FX-series - for a long time ... intel knew better then to pour money into an engineering failure as there was no *hope* for it ... barring a 'miracle'

how basic is that?
:roll:

unless you are in denial?

None of what you said makes Netburst an "engineering failure". It didn't reach higher clock speeds because of process limits, not architectural limits.

 

[apoppin]

Originally posted by: zsdersw

Originally posted by: apoppin
at least the 'pot' can use Mr Kettle's own words

No, you can't.

to show he also agrees that the P4 is an 'engineering failure' ... he just refuses to acknowledge his own words

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.

intel's best engineers failed to *solve* the thermal limit "problem" of P4 and the project was canned because P4 was not competitive with FX-series - for a long time ... intel knew better then to pour money into an engineering failure as there was no *hope* for it ... barring a 'miracle'

how basic is that?
:roll:

unless you are in denial?

None of what you said makes Netburst an "engineering failure". It didn't reach higher clock speeds because of process limits, not architectural limits.

you fanboys are a lot of fun ... better 'n video



apoppin brings out the big[er] guns
:evil:

Marketing blunders of our time

Intel - the Pentium 4 will hit 10GHz
The Pentium processor has been one of the biggest names of the last decade. It launched in 1993, and the name came from the fact that it was the fifth-generation x86 architecture chip from Intel.

The chip had several successors, handily named the Pentium 2, Pentium 3 and Pentium 4. All scaled the amount of transistors, clock speed and cache included on the chip, theoretically scaling the power and speed at the same time.

The Pentium 4 was released in November 2000, with the first revision codenamed 'Willamette'. It employed a microarchitecture called NetBurst, which was designed to provide the fastest, most efficient processing for the programmes of the time. Included in NetBurst was a rapid execution engine, where the ALUs in the CPU actually operated at double the frequency of the core clock, and Hyper Pipelined technology, which was the extra-long number of stages in the processing pipeline - Willamette had 20 stages whilst Prescott, the last revision of the Pentium 4, had an amazing 31.

The Pentium 4 was heralded as a chip which would be at the heart of computers for a decade, scaling up to 10GHz by following Moore's Law.

In an article for PC World, Intel's George Alfs was quoted saying:

"Intel's P4 at 2 GHz is the world's highest performance desktop processor... P4's NetBurst architecture has the ability to scale to 10 GHz in its lifetime."




Articles all over the net quoted chip guru Albert Yu as suggesting that the P4 would scale up to 10GHz and beyond through the lifetime of the chip.

Even better were the articles that attempt to compare the Pentium 4 to the Pentium 3. We now know, of course, that the P4's monstrously long pipeline was responsible for terrible heat problems and some dire performance. But back at the launch in 2000, Intel was playing down the drop in efficiency that the P4 represented - even though, as we know now, Intel has reverted back to a Pentium 3-like architecture with Core Duo. The quotes comparing the two are amusing:

"A longer pipeline makes for higher clock speed. But does it sacrifice performance? The answer is yes and no.

Asked if a Pentium III would outperform a Pentium 4 at the same clock speed, Intel's Albert Yu said, "It's technically correct, but it's artificial."

Forthcoming Pentium III chips based on 0.13-micron process -- which are expected to reach speeds as high as 1.4GHz -- would outperform the Pentium 4 chips due out this fall on 0.18-micron technology. But Yu said that was an apples-to-oranges comparison."

10.20GHz Intel Nehalem slated for 2005

in '04 even these morons got it right
The Roadmap to Recovery: Part I 10GHz was dreaming the impossible dream

Prestonia enhances the Xeon line in three ways, according to Lisa Hambrick, director of enterprise processor marketing at Intel. First, by switching to the Pentium 4 architecture, Intel can drastically boost the clock speed. The old server Xeon topped out at 1.4GHz. The new one debuts at 1.8GHz, 2GHz and 2.2GHz, and will eventually pass 10GHzText

More?


http://www.infosatellite.com/news/2001/08/j0801intel35ghz.html

"Yesterday, we showed a 2-GHz processor. Today, we showed a 3.5-GHz processor. A 4 GHz processor is on the horizon," said Paul Otellini in a speak at the conference, Executive Vice President and general Manager of Intel´s Architecture Group. "We're convinced that we can scale the Pentium 4 to 10-GHz," he added.

http://en.wikipedia.org/wiki/Pentium_4

Upon release, the Prescott turned out to generate approximately 40% more heat clock-for-clock than the Northwood, and almost every review of it was negative, earning it the soubriquet PresHot. A shift in socket type (from Socket 478 to LGA775) was expected to reduce the heat to more acceptable levels, but in fact proved to have the opposite effect, with power requirements increasing by a further 10%. However, the LGA775 reference cooler and mounting system were somewhat better designs, so average temperatures were slightly lowered. Subsequent revisions to the processor by Intel engineers were expected to reduce average temperatures, but this never happened outside of the lowest speed grades. Prescott Pentium 4s were given the product codes 80546 (Socket 478) and 80547 (LGA775).

Finally, the thermal problems were so severe, Intel decided to abandon the Prescott architecture altogether, and attempts to roll out a 4 GHz part were abandoned, as a waste of internal resources. Also of concern was the fact that a review showed that in games, it took a 5.2 GHz Prescott core to soundly beat the performance of an Athlon FX-55 that clocked at 2.6 GHz [5]. Considering Intel boasted at launch the Pentium 4 architecture was intended to support up to 10 GHz operation, this can be seen as one of the most significant, certainly most public, engineering shortfalls in Intel?s history. This also meant that while Northwood ultimately achieved clockspeeds 70% higher than Willamette did, Prescott only managed a 12% rise over Northwood.

argue with *them* ^^^ >> the rest of the tech world

:roll:

up da'Nile in a paper boat smoking a cigarette

 

[Jadow]

Think of all the millions of P4's that are sitting in offices around the world. It definitely wasn't a financial failure for Intel.

 

[zsdersw]

Originally posted by: apoppin
apoppin brings out the big[er] guns

Your "big guns" did nothing but demonstrate what I've been saying for quite a while: marketing failed and made mistakes.. not engineering and not the architecture.

Anyone who regards a marketing claim that didn't come true as a demonstration of engineering failure or shortcoming is a fool.

 

[coldpower27]

Originally posted by: apoppin
after all *that* long-winded spiel, you really agree the P4 IS an engineering failure

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.

:roll:

some much for the "validity" of your argument ... self-contradictions are usually a clue ... "thermally limited" with NO solution = an 'engineering failure'

in ANY language and with an agreed-upon universal *dictionary* definition of FAILURE
:thumbsdown:

spin it all you want

you have no grasp of it at all ... i don't want to even bother to address the rest of your fantasy ... the main point is that the intel Engineers could not *fix* the P4's many problems including your self-admitted thermal problems ... so it was ABANDONED as a 'failure'.

Since my definition of engineering failure is different to yours I completely disagree, and like I have said before being thermally limited is not an engineering failure but as a result of process problems on that node. There is no contradiction here, you are seeing something which is simply non-existant. The goals have simply changed now, and NetBurst was never designed to meet the new goals, so changing the architecture here is anything but failure because it is not realistic to use NetBurst for something it was never intended to do.

 

[imported_Truenofan]

i would believe intel did have an engineering failure, it failed to compete in real tests, to believe it wasnt is nonsense. im neither intel nore amd fanboy, but the p4 i knew from the start, was a joke. as soon as i had my a xp, i loved it, it beat anyones computer around me easily, then the athlon 64 3500, another fast cpu, and now i have my intel c2d, and i love it. fastest thing i've set my hands on. why dont some people admit that it didnt achive what intel expected at all.

 

[apoppin]

Originally posted by: coldpower27

Originally posted by: apoppin
after all *that* long-winded spiel, you really agree the P4 IS an engineering failure

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.

:roll:

some much for the "validity" of your argument ... self-contradictions are usually a clue ... "thermally limited" with NO solution = an 'engineering failure'

in ANY language and with an agreed-upon universal *dictionary* definition of FAILURE
:thumbsdown:

spin it all you want

you have no grasp of it at all ... i don't want to even bother to address the rest of your fantasy ... the main point is that the intel Engineers could not *fix* the P4's many problems including your self-admitted thermal problems ... so it was ABANDONED as a 'failure'.

Since my definition of engineering failure is different to yours I completely disagree, and like I have said before being thermally limited is not an engineering failure but as a result of process problems on that node. There is no contradiction here, you are seeing something which is simply non-existant. The goals have simply changed now, and NetBurst was never designed to meet the new goals, so changing the architecture here is anything but failure because it is not realistic to use NetBurst for something it was never intended to do.

as a result of process problems on that node
more engineering failure ... they could NOT *fix* it

and thanks for the Spin
:roll:

fortunately 90% of the informed Tech world knows better
- those who can agree with the dictionary definition of what "failure" means ... not some fantasy speculation of reinvented history

you have only quoted yourself

find *something* to back up what you say - besides other fanboys
:Q

 

[THUGSROOK]

i think some ppl forget all the P4 fun we had

northwood was far from a failure, prescott was a serious failure tho.

 

[zsdersw]

Originally posted by: Truenofan
i would believe intel did have an engineering failure, it failed to compete in real tests, to believe it wasnt is nonsense.

Umm.. no. For a significant majority of Netburst's life, it was at least competitive to AMD's offerings, winning some benchmarks and losing others.

 

[zsdersw]

Originally posted by: apoppin
and thanks for the Spin

Actually, it's us who should be thanking you for the spin.

 

[apoppin]

Originally posted by: zsdersw

Originally posted by: apoppin
and thanks for the Spin

Actually, it's us who should be thanking you for the spin.

l posted facts, substantiated links and history ...

otoh, you posted ... your personal opinion with *zero* ... zip ... nada ... NOTHING to back you up

ridiculous claims of the P4's superiority:

Originally posted by: zsdersw
Umm.. no. For a significant majority of Netburst's life, it was at least competitive to AMD's offerings, winning some benchmarks and losing others.

significant? ... maybe a year

it was a *disaster* on launch failing to compete with the PIII ... it didn't get 'competitive' till the NW b and c series then it FLOPPED with Prescott again ... a hot fire-breathing monster with crap performance they couldn't fix

P4 finished as a *loser* against FX at half the price

a dud

a flop

an engineering failure

intel's marketing and Dell's undying support [for cheap CPU prices] kept them going till they had a *winner* again

only a fanboy holds on to a fantasy long after the company he worships completely demolishes it

intel would like to forget the P4 debacle ... it is an embarrassment to them

:roll:

 

[apoppin]

Originally posted by: THUGSROOK
i think some ppl forget all the P4 fun we had

northwood was far from a failure, prescott was a serious failure tho.

NW was *wonderful* .. the pinnacle of the P4 architecture and .. yes .. engineering

i did *not* forget the fun we had or the FACT that you and i ... both with ANCIENT agp intel rigs ... well over 3 years old each ... are *Still competitive* with the latest midrange [plus] PCIe rigs in gaming ... nevermind office uses.

However, that does not *excuse* the fact that intel's engineers intended "much more" than 3.73Ghz with the last 1066FSB Gallatin/NW CPU

they ran into "issues" that prevented them from reaching 4Ghz ... forget 10ghz as projected in '05.

instead of attempting to shrink NW's core, they unwisely decided on Prescott ... it was supposed to run cooler

but they failed ... instead, they kept getting *misses* in their 'predictions' in the ever-lengthening pipelines of Prescott ... intel's engineers realized it *could* take many years to find and identify and fix the myriad problems plaguing the *process*

that IS an engineering failure ... intel cut their losses, tossed P4 and NetBurst in the trash and went back to perfecting PIII-M type CPUs with success.

the above is history ... as told by an intel fanboy ... since my first CPU in my Atari800xl [MOS 6502C ... i think it was 1.77Mhz and i did not try to OC it] ... then a long gap of 'nothing' - all my rigs have had 'intel inside'

.... purely coincidently ... i would not have minded ... or would mind even now, AMD
[btw, i got to try my first nvidia GPU - 7800GS OC .. since my first GeForce SDR ... only last month - nice! ... after a run of all ATi ... weird ]


btw, TR ... what is the NB fan to get for the IC7? ... mine is finally starting to annoy me

 

[kobymu]

Now you shouldn?t have said THAT ?

However, that does not *excuse* the fact that intel's engineers intended "much more" than 3.73Ghz with the last 1066FSB Gallatin/NW CPU

It wasn?t the engineers intent, the intent was of the management/marketing, the engineers where just being good little foot soldiers how wore sent in to the battle field without enough training (time) and equipment (money)...

The engineers did took them pretty far for the time, considering the fact they pushed the envelope (equal to going against an unknown enemy of unknown size with unknown capabilities) as far as they probably could under the restraints of time and money they had, Intel is a business, not an academic research company with unlimited money and no time restraints.

Yes Intel may very well by a 800-pound gorilla, but even a 800-pound gorilla cannot just start venture into the unknown for more then 5 years in a row and have a list comprised only of successes while maintaining a profitable bottom line! The unknown isn?t a forgiving place, not even to an 800-pound gorilla.

 

[apoppin]

of course the intel engineers "intended" more ...

*who else* advised their superiors of progress and whether they *thought* they could get more Ghz with Prescott or with NW?

the management took the *advice* and opinion of their engineers and the best [in their opinion] gamble to get more Mhz in the speed race with AMD

i bet some heads rolled over that one at intel
:Q

*someone* took the blame for P4's failure ... i never followed the political aftermath

 

[myocardia]

Originally posted by: 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.

Do you honestly believe this crap that's spewing from your lips/fingertips? Wait, I'll put on my hipwaders. The definitions of both of these words have been the same for many hundreds of years: #1 engineering and #2 failure. Does your thesaurus, assuming you own one, say that failure=success? Because, from what I keep seeing you say, it sure does seem like it. Oh, and where exactly did you think that the people in the marketing dept. got the idea that Netburst was capable of ~10 Ghz, from a secretary?:laugh:

 

[myocardia]

Originally posted by: coldpower27
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.

If you're going to be a part of a society, and/or speak a language, you can not just make up definitions as you see fit. See the above post for the definitions of both "engineering" and "failure".

 

[imported_Truenofan]

they stated continually that the p4 was going to hit 10ghz, i even remeber when it was originally released, it souned impressive at the time. but then again i was one of those "speed is what counts" at the time. either the marketing dept. thought the idea up on they're own, or the engineer's stated that it might be possible to hit the 10ghz with the p4.

 

[natto fire]

Originally posted by: myocardia

Originally posted by: coldpower27
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.

If you're going to be a part of a society, and/or speak a language, you can not just make up definitions as you see fit. See the above post for the definitions of both "engineering" and "failure".

LOL, perhaps too many people on here see politicians as role models?

Yes, they still made boatloads of money from Netburst chips, but they engaged in some ethically questionable business tactics to do so. When more and more OEMs started shipping AMDs despite Intel doing everything in their power to squash them in this market, they probably really started to take notice. (At this time the enthusiast market already was mostly on AMD)

Do people really think hardware companies let their marketing departments decide on engineering goals? Is that where all those chip roadmaps come from? I always thought that an engineer or two might help make those. Even if this were the case, then the engineers failed to tell the marketing department that it might take significant time, if even possible, to reach 10GHz.

 

[zsdersw]

Originally posted by: myocardia
Do you honestly believe this crap that's spewing from your lips/fingertips? Wait, I'll put on my hipwaders. The definitions of both of these words have been the same for many hundreds of years: #1 engineering and #2 failure. Does your thesaurus, assuming you own one, say that failure=success? Because, from what I keep seeing you say, it sure does seem like it. Oh, and where exactly did you think that the people in the marketing dept. got the idea that Netburst was capable of ~10 Ghz, from a secretary?:laugh:

Try to keep up with the conversation, will ya? This has nothing to do with dictionary definitions of "engineering" and "failure". It does have everything to do with how the specific meaning of those words applies in this situation. Given the vague-ness of the phrase "engineering failure", there is ample room for the debate we've been having, in spite of your pithy (and totally irrelevant) dictionary-thumping.

"10GHz" sounds perfect from a marketing perspective.. and like *way* more than a pipe dream to an engineer. Now who do you think was most likely to have pushed that "10GHz" idea? A marketer who sees only dollar signs and market share bar graphs.. or an engineer who sees the transistors and circuits that make it all happen? Engineers are generally not shoot-the-moon types; they don't talk about things matter-of-factly that are too far beyond what's already been done.