As an engineer I have to disagree with you here Taltamir...engineering failure invokes a rather specific implication regarding the product which was engineered.
Just because a loose group of educationally challenged folks who have internet access and forum membership decide to apply liberal interpretations to the intent of the words "engineering failure" does not mean any one of those interpretations (regardless how passionately they are prosed) are correct.
This thread is really pathetic IMHO, you guys are arguing over something that truly none of you appear to have the competence/education/experience to be an authority to argue about. Dmens had this spot on in one of his earlier posts.
This thread is a waste of effort and time for everyone here, you may as well be arguing over religion versus atheism...even if you were an AMD cpu design engineer you wouldn't possibly have access to all the necessary info you would need regarding the the dynamic internal decision process that was going on with the netburst technology at the engineering level of decision making at Intel to formulate a credible opinion on the thread title's question.
I can only imagine the Ford and GM forums are full of equally passionate carheads who have no idea of what all goes into the design/manufacturing cycle of an automobile but they are equally confident in their abilities at single-handedly enumerating why and how the ceo of the company led to its demise.
The monday morning quarterback stereo-type exists for a reason. Do you folks really fancy yourselves capable of second-guessing years and years worth of decisions made by hundreds and thousands of qualified (education and experience) engineers and product managers at Intel (or AMD for all those Phailnom-minded posters)? Give me a break.
- Sep 28, 2005
- 21,026
- 3,497
- 126
Im sorry, a success in my book means something that worked and ur improving.
Failure means u need to take a step back and re evaluate.
So when your on P4 tech for years, and u take a step back to P3 what is that called in your definition?
to me its a failure.
Why because u need to take a step back to the older tech and mature it, because your current tech blows.
That is what happened to the end of P4.
GET IT STRAIGHT PEOPLE... C2D is not a MATURE P4... ITS A P3!
If intel got it straight it would of gone... P1 -> P2 -> P3 -> C2D (NOT P4)
Instead we did this...
P1 -> P2 -> P3 -> P4 -> P3 -> C2D
U guys still want to call that a non failure?
U repeat something when u failed... ummm P3 got repeated on YONAH.
YONAH is the first C2D.
If u were to ask intel directly... they would probably smile it and avoid the question and say.. well we got i7's.
Even they know its a failure... they took a lot of hit when K8's started eatting market shares.
Arkaign
Lifer
- Oct 27, 2006
- 20,736
- 1,379
- 126
C2D is not even close to a P3, although it is a descendant of the basic design (that started with PPro, then P2, then P3).
"The Pentium M represented a new and radical departure for Intel, as it was not a low-power version of the desktop-oriented Pentium 4, but instead a heavily modified version of the Pentium III Tualatin design (itself based on the Pentium Pro core design)"
"Yonah was the code name for (the core of) Intel's first generation of 65 nm process mobile microprocessors, based on the Banias/Dothan-core Pentium M microarchitecture. SIMD performance has been improved through the addition of SSE3 instructions and improvements to SSE and SSE2 implementations, while integer performance decreased slightly due to higher latency cache. Additionally, Yonah includes support for the NX bit."
These developments are HIGHLY evolutionary in nature. Calling C2D a Pentium 3 is something akin to calling i7 a P4.
The P4 continually improved until the speedbump of Prescott, then they corrected things with Cedar Mill, but it was too late, AMD had caught up and passed them with AMD64. The exact same thing happened to AMD with their Socket A chips, they led early, then were dead even with the Northwood A and B series, then lagged behind the Northwood C's, which got worse as they failed to ramp up to compete with the 3.2C. Thoroughbred A was their speedbump, and it cost them enough time to fall behind in the race. Simply because Athlon XP was replaced does not make it a failure.
"By the time of Barton's release, the Northwood Pentium 4 had become more than competitive with AMD's processors.[22] Unfortunately, due to the architecture of AMD's processor caches, an L2 cache increase to 512 KiB did not have nearly the same impact as it did to Intel's line. Only an increase of several percent was gained in per-clock performance.[21] The PR rating became somewhat inaccurate because some Barton models with lower clock rate weren't consistently outperforming their higher-clocked Thoroughbred predecessors with lower ratings.[22]
The other improvement, a higher 400 MT/s bus clock, helped Barton gain some more efficiency. However, it was clear by this time that Intel's quad-pumped bus was scaling well above AMD's double-pumped EV6 bus. The 800 MT/s Pentium 4 bus was well out of Athlon's reach. In order to reach the same bandwidth levels, the Athlon bus would have to be clocked at levels simply unreachable.[21]
The K7 architecture had scaled to its limit. Maintaining performance equivalence with Intel's improving processors would require a significant redesign."
You'd have to be a moron to call the K7 a failure in any way, but it did end up needing replacement. The AMD64 chips, although based on many K7 functions, were revolutionary in design, particularly for multi-core and memory access.
The K7 led for a good amount of time, and so did the P4 ..
BOTH WERE REPLACED DUE TO THE COMPETITION RELEASING SUPERIOR PRODUCTS.
This is what happens in a competitive business. A failure is something that never meets the goal of offering a competitive product to gain your company the desired sales. As said before : things like i740, RDRAM, K5, etc.
zsdersw
Lifer
- Oct 29, 2003
- 10,505
- 2
- 0
Originally posted by: myocardia
You know, I've noticed over the years that only children and helmet-wearing short bus riders start name calling when proven wrong. BTW, how low of an IQ does it require to think the best selling CPU of all time was a marketing failure?The Pentium 4 was the antithesis of a marketing failure. Who, besides you, didn't know this?
Perhaps you missed my post from the top of this page:
That's not the marketing I'm referring to. I'm referring to the people in Intel who were stuck on more and more clock speed at any cost to keep up in the hertz race because that's how things were marketed.
The decision by marketing and management to push clock speed above all else is what drove the development of the P4; the best way the engineers could come up with at the time to yield faster clocks: very deep pipeline (31-stage in the Prescott, IIRC).
But neither of these occurred. It consistently met engineering standards. Xbox 360 is an example of engineering failure, Pentium 4 is not.
Also this thread is from February 2007
taltamir
Lifer
- Mar 21, 2004
- 13,576
- 6
- 76
engineering failure by which definition? there are records of the xbox engineers saying EXACTLY HOW and WHY the xbox is going to fail, every time management decided to "cut costs" DESPITE what the engineers said, and every time it failed as the engineers predicted.
Did the engineers FAIL, or is the xbox engineered badly DESPITE the engineers because they did not have the final say?
Does engineering fail mean:
1. Engineers failed to design it properly.
2. Engineers designed it properly, but management made changes to it and caused it to become a failed design.
I have personally witnessed 2 on multiple occasions. Heck I have had it DONE to me... where a clueless boss told me what parts to use, how to connect them, and exactly how to do my job wrong... and would not listen when I explained why it is a bad idea...
Did the engineers FAIL, or is the xbox engineered badly DESPITE the engineers because they did not have the final say?
Does engineering fail mean:
1. Engineers failed to design it properly.
2. Engineers designed it properly, but management made changes to it and caused it to become a failed design.
I have personally witnessed 2 on multiple occasions. Heck I have had it DONE to me... where a clueless boss told me what parts to use, how to connect them, and exactly how to do my job wrong... and would not listen when I explained why it is a bad idea...
Engineering failure usually means something doesn't work the way it was intended to work.
That might seem like a sketchy definition but I'll give you an example that I heard from a chemistry professor trying to explain quality control and quality assurance. Suppose I sell you a parachute and I say "This parachute will only work 1 time." You use it 1 time and it works. You try to use it a second time, it fails, and you fall to your death. Did I sell you a quality product? The answer is actually yes; it did exactly what I said it would do. It may not do what you wanted it to do, but it did what I said it would do. This is how we define things like "quality" or "engineering failure".
When it comes to the P4, it did exactly what Intel said it would do. They say it will work perfect at stock speed and stock voltage for this time frame and get the same calculation every time. This is the standard they put out, and it met that standard. If would only be an engineering failure if they promised more than it delivered. An example of this kind of failure would be when I bought a factory overclocked video card and it couldn't pass ATI Tool's artifact test.
That might seem like a sketchy definition but I'll give you an example that I heard from a chemistry professor trying to explain quality control and quality assurance. Suppose I sell you a parachute and I say "This parachute will only work 1 time." You use it 1 time and it works. You try to use it a second time, it fails, and you fall to your death. Did I sell you a quality product? The answer is actually yes; it did exactly what I said it would do. It may not do what you wanted it to do, but it did what I said it would do. This is how we define things like "quality" or "engineering failure".
When it comes to the P4, it did exactly what Intel said it would do. They say it will work perfect at stock speed and stock voltage for this time frame and get the same calculation every time. This is the standard they put out, and it met that standard. If would only be an engineering failure if they promised more than it delivered. An example of this kind of failure would be when I bought a factory overclocked video card and it couldn't pass ATI Tool's artifact test.
DrMrLordX
Lifer
- Apr 27, 2000
- 22,560
- 12,426
- 136
Originally posted by: taltamir
What we are having here is mostly a communication failure
http://www.youtube.com/watch?v=AShBoF1FPSE
'nuff said.
Yep.
Originally posted by: taltamir
ok... that is different than both the definition I figured for it, and I have seen people use throughout this thread.
What we are having here is mostly a communication failure
I'll quote myself (3rd post to this very thread) from some 2.5 yrs ago:
It was a failure, because Intel is a multi-billion dollar company that was caught sitting on their asses !! If you don't see that you have blinders on !!
myocardia
Diamond Member
- Jun 21, 2003
- 9,291
- 30
- 91
That's because you never made this statement in the post I quoted, or in any post prior, at least that I saw. If you had, I wouldn't have disagreed with you in the first place.
What you describe here is very far from an engineering failure. The engineers in this instance did everything they were supposed to do, including correctly concluding how the parts would fail, if changes were made to it. This failure was quite obviously (at least in my mind) 100% management's fault, making it a management failure, not an engineering failure.
myocardia
Diamond Member
- Jun 21, 2003
- 9,291
- 30
- 91
Over the years (seriously ), I keep meaning to bring this up in this thread, and keep forgetting or getting distracted. The main reason that I believe the P4 was an engineering failure has absolutely nothing to do with it reaching far less than half of its predicted final speed. The reason that I consider it a failure is because out of the 8 or 9 iterations of the product, a total of two of them were considered by the people who count as being products clearly worth buying, worth owning, worth recommending. The people who count, IMO, are professional reviewers and computer afficionados like us, whose opinions convince most people on the planet which hardware is better (<<--both of us, not just the afficionados). That's a 25% or less success rate. The only place that I've ever heard of a 25% success rate not being horrendous is professional baseball, and even there a .250 batting average is nothing to brag about.
edit: The two iterations of the P4 to which I referred were the Northwood C and the Cedar Mill.
), I keep meaning to bring this up in this thread, and keep forgetting or getting distracted. The main reason that I believe the P4 was an engineering failure has absolutely nothing to do with it reaching far less than half of its predicted final speed. The reason that I consider it a failure is because out of the 8 or 9 iterations of the product, a total of two of them were considered by the people who count as being products clearly worth buying, worth owning, worth recommending. The people who count, IMO, are professional reviewers and computer afficionados like us, whose opinions convince most people on the planet which hardware is better (<<--both of us, not just the afficionados). That's a 25% or less success rate. The only place that I've ever heard of a 25% success rate not being horrendous is professional baseball, and even there a .250 batting average is nothing to brag about.edit: The two iterations of the P4 to which I referred were the Northwood C and the Cedar Mill.
(June 2003)
gaming 1 - P4 wins
gaming 2 - P4 wins
gaming 3 - P4 wins
gaming 4 - P4 wins
gaming 5 - P4 wins
gaming 6 - P4 wins
video encoding - P4 wins
3D rendering 1 - P4 wins
3D rendering 2 - P4 wins
final words
A year later (June 2004), AMD's lovely Socket 939 comes out. *dramatic music*
video encoding - tie
3D rendering - tie
compiling source code - Athlon wins
AMD catches up! Yay!
AMD eventually had domination while Intel was dicking around with Prescott (October 2004), but it didn't last very long. Intel releases the Pentium D (April 2005).
encoding - Pentium wins
rendering, multitasking - Pentium wins
Not to be beaten so easily, AMD releases their own dual cores (May 2005).
encoding, gaming, rendering - Athlon wins
A few months later, Intel releases the Core Duo which is roughly equal to AMD's processors. (December 2005)
(July 2006) Intel releases the Core 2 Duo which beats the Athlon across the board.
rendering 1, rendering 2, gaming 1, gaming 2, gaming 3 - Core 2 Duo wins
final thoughts
So as you can see, AMD never had a strong hold on anything. The golden Athlon age was very similar to current Phenom II situation.
myocardia
Diamond Member
- Jun 21, 2003
- 9,291
- 30
- 91
Anyone can cherrypick results, Shawn. Or were all P4's really slower than all P3's and Atlon XP's? See how easy that was?
edit: BTW, why would you be linking C2D benchmarks in a thread solely about Pentium 4's?
edit: BTW, why would you be linking C2D benchmarks in a thread solely about Pentium 4's?
Originally posted by: myocardia
Anyone can cherrypick results, Shawn. Or were all P4's really slower than all P3's and Atlon XP's? See how easy that was?
edit: BTW, why would you be linking C2D benchmarks in a thread solely about Pentium 4's?
Anandtech's article from November 2000 (same month as your link)
games - P4 wins
more games - tie
more games - Athlon wins
OpenGL - P4 wins 1, loses 2
more OpenGL - P4 wins 1, loses 1, ties 1
What a horrible piece of shit processor! I mean look at it. It can only tie AMD's best processor at the time. The engineers at Intel are so dumb that they should just kill themselves.
A few months later (April 2001), the battle is between AMD's Athlon Thunderbird and the newly released 1.7ghz P4.
gaming - P4 wins 1, loses 1
more gaming - P4 wins 1, loses 1
encoding - Athlon wins
What garbage. Those Intel engineers need to buy a dictionary so they can learn what seppuku is.
A few months later (August 2001), the P4 hits 2ghz
encoding - Athlon wins
games - P4 wins 2, loses 1
Fast forward to January 2002, first of the P4 Northwood
CAD - P4 wins 1, loses 1
more CAD - Athlon wins
3D rendering - tie
more 3D rendering - still a tie
encoding - tie
gaming - P4 wins 1, loses 1, ties 1
Anand's final words
(I bought an Athlon 1700+ at this time because it was cheaper as Anand states)
By the end of the year (November 2002), hyperthreading came out and the P4 was in the lead. Athlon still had some fight in gaming but everything else mostly favored the P4.
first page of benchmarks, starting with encoding
So tell me, where is this Athlon domination I hear about? We've gone through the entire sequence of events from 2000 to 2006 and we still haven't seen it. The infamous P4 Willamette was the same speed as the Athlon Thunderbird then the P4 Northwood was the same speed as the Palamino and Thoroughbred up until the P4c. When the Pentium 4c hit, the P4 was better than the Barton in every test, the P4 could overclock higher, and the P4 ran a lot cooler. The only reason any of us bought an Athlon was because they were cheaper. The only time the P4 was a clear loser across the board was the Prescott, and that didn't last very long.
I wonder where everyone was during the P4 era. If people are posting stuff about how slow the P4 is, I can't help but think they have absolutely no idea what they're talking about. Anandtech's old articles are all available and you don't need to pay for them, so why hasn't anyone bothered to look at any of them? I would post Tom's Hardware articles as well to confirm everything Anand said over the past 9 years but people will just write that off because "Tom is an Intel fanboy" even though his results are often similar to everyone else's results.
myocardia
Diamond Member
- Jun 21, 2003
- 9,291
- 30
- 91
Originally posted by: ShawnD1
Originally posted by: myocardia
Anyone can cherrypick results, Shawn. Or were all P4's really slower than all P3's and Atlon XP's? See how easy that was?
edit: BTW, why would you be linking C2D benchmarks in a thread solely about Pentium 4's?
So tell me, where is this Athlon domination I hear about? We've gone through the entire sequence of events from 2000 to 2006 and we still haven't seen it. The infamous P4 Willamette was the same speed as the Athlon Thunderbird then the P4 Northwood was the same speed as the Palamino and Thoroughbred up until the P4c. When the Pentium 4c hit, the P4 was better than the Barton in every test, the P4 could overclock higher, and the P4 ran a lot cooler. The only reason any of us bought an Athlon was because they were cheaper. The only time the P4 was a clear loser across the board was the Prescott, and that didn't last very long.
Reading comprehension problems, huh? Don't worry, that's a very common problem around here, it seems. Try reading what you quote, though. It makes it seem much more like you might know what you're talking about.
IntelUser2000
Elite Member
- Oct 14, 2003
- 8,686
- 3,786
- 136
Uhh, sure P4 did have its moments of success. I don't count Cedarmill as they were about as successful as the Phenom IIs are now. Competitive with pricing, but nowhere near the performance. The only time was during the Northwood B and C days. B allowed to catch up, C allowed to exceed it.
Otherwise, there was no time where P4 didn't "fail" whether it was marketing or engineering. On the Pentium D days AMD briefly had higher marketshare than Intel on the retail desktop market!
P4 Willamette Nov 2000-Introduced with substantial delay(6+ months), had a problem with BIOS which was fixed merely a day before release, didn't perform to par, didn't take the performance lead. Costed as much as the QX9770 for supposedly mainstream versions. Used Rambus and broke compatability completely.
P4 Northwood Jan 2002-Lower power, higher clock and minor per clock improvements allow competitiveness with Athlon
Northwood HTT-This is about the time Intel eeks a performance lead
Northwood C-The combination of high clock speeds, HTT, faster FSB and superior platform lacking Rambus, along with AMD slowing down has Intel taking the lead, briefly.
From then its all downhill. The EE versions which merely caught up with the high end Athlons, the crappy Celerons which were compared to Via(figuratively speaking as the gap compared to Semprons was too large), dual core chips that reached 150W+ in power which had low end AthlonXP performance, significant delays. Next big revision chip delivering nothing(AKA Prescott). On top of the expensive prices.
Combination of falsely-driven marketing with failed execution.
Otherwise, there was no time where P4 didn't "fail" whether it was marketing or engineering. On the Pentium D days AMD briefly had higher marketshare than Intel on the retail desktop market!
P4 Willamette Nov 2000-Introduced with substantial delay(6+ months), had a problem with BIOS which was fixed merely a day before release, didn't perform to par, didn't take the performance lead. Costed as much as the QX9770 for supposedly mainstream versions. Used Rambus and broke compatability completely.
P4 Northwood Jan 2002-Lower power, higher clock and minor per clock improvements allow competitiveness with Athlon
Northwood HTT-This is about the time Intel eeks a performance lead
Northwood C-The combination of high clock speeds, HTT, faster FSB and superior platform lacking Rambus, along with AMD slowing down has Intel taking the lead, briefly.
From then its all downhill. The EE versions which merely caught up with the high end Athlons, the crappy Celerons which were compared to Via(figuratively speaking as the gap compared to Semprons was too large), dual core chips that reached 150W+ in power which had low end AthlonXP performance, significant delays. Next big revision chip delivering nothing(AKA Prescott). On top of the expensive prices.
Combination of falsely-driven marketing with failed execution.
IntelUser2000
Elite Member
- Oct 14, 2003
- 8,686
- 3,786
- 136
myocardia
Diamond Member
- Jun 21, 2003
- 9,291
- 30
- 91
You really don't think the Cedar Mill was a nice chip, which would have been much more successful, had it not been for how well Intel had pre-promoted the C2D's?
The Northwood C's were dominant. They were just too expensive, compared to their competition.
edit:
Please don't remind me of how much computer parts used to cost! I paid a lot more for a 600 Mhz Athlon in '98 (or was it early '99?) than I did for my Q6600 right after the release of the G0's. BTW, my father paid ~$4,000 for my first computer.
Oh, and I can't remember what year it was now, but I remember paying $200 for 2MB of RAM, also.
taltamir
Lifer
- Mar 21, 2004
- 13,576
- 6
- 76
Originally posted by: myocardia
Over the years (seriously
edit: The two iterations of the P4 to which I referred were the Northwood C and the Cedar Mill.
and government...
taltamir
Lifer
- Mar 21, 2004
- 13,576
- 6
- 76
Originally posted by: Idontcare
Yep.
Originally posted by: taltamir
ok... that is different than both the definition I figured for it, and I have seen people use throughout this thread.
What we are having here is mostly a communication failure
I'll quote myself (3rd post to this very thread) from some 2.5 yrs ago:
thank you for clarifying... can you tell me than what would be the term for the two definitions I gave?
management failure? product failure? marketing failure?
Northwood C were awesome chips at the time, had a 3.2ghz one at the time. Athlon XPs couldn't touch them (and they tended to cook themselves)
I would have to say even single core Athlon 64s weren't much of an upgrade, sure they were faster its just they were horrible in comparison under heavy multitasking. (I had both a 3200+ venice and p4c 3.2 side by side at the time, chose to stick with the athlon 64 though)
I would have to say even single core Athlon 64s weren't much of an upgrade, sure they were faster its just they were horrible in comparison under heavy multitasking. (I had both a 3200+ venice and p4c 3.2 side by side at the time, chose to stick with the athlon 64 though)
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 24K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes Discussion Threads
- Started by Tigerick
- Replies: 20K
-
-
-
Facebook
Twitter