Was the P4 an 'engineering failure'?

[SickBeast]

This is to settle a debate.

I want to hear opinions on why the P4 was a failure, or why it wasn't.

Personally I think it was, and it only survived because of intel's marketing and brand recognition.

:beer:

 

[Agentbolt]

Not a diamond member, but it's hard to argue against AMD increasing their marketshare severalfold during the P4 era. Wouldn't have happened if the P4 had been able to run faster for the price.

 

[Idontcare]

I liked my P4, it did the job for me. No complaints.

As to whether it was an engineering failure...not sure what profession you harken to, but I am an engineer and there is nothing about the P4 that failed.

It did exactly as engineered to do. If it failed from engineering then it would not function.

Maybe you don't understand what engineering is or what failure is?

Your thread title is akin to stating mathematics is a failure because 2+2 = 4 when what you personally wanted was 1+1 = 2. Yes 2+2=4 and yes 1+1=2, neither is a failure and neither is bad mathematics. But if you personally wanted "2" then don't add 2+2 to get 4 and then claim mathematics is a failure.

The P4 is exactly as it was engineered to be. Whether you wanted a P4 or not is probably the only question you should ask yourself.

(No I do noyt and have not ever worked for Intel, but I am an engineer in the semiconductor industry)

 

[3chordcharlie]

It was an engineering failure in terms of not doing what intel set out for it to do.

It was a product failure in terms of being beaten by Athlons for a large part of its run. But it was never beaten badly, and at certain times was the best product available.

The P4 turned out to be a poor design, but it wasn't an overall failure - it was a reasonable product.

I would suggest that netburst-based celerons did more to lose intel market share than P4 did. They were truly awful products. Possibly the worst cpu offered since the original no-cache celerons, and a terrible decision considering AMD's ability to offer low-cost Athlons, durons, and later semprons that absolutely whipped the celeron. If there was an engineering failure of the P4, it was complete inability to strip cache (saving a lot of manufacturing cost) and still maintain usable performance.

 

[coldpower27]

Originally posted by: 3chordcharlie
It was an engineering failure in terms of not doing what intel set out for it to do.

It was a product failure in terms of being beaten by Athlons for a large part of its run. But it was never beaten badly, and at certain times was the best product available.

The P4 turned out to be a poor design, but it wasn't an overall failure - it was a reasonable product.

I would suggest that netburst-based celerons did more to lose intel market share than P4 did. They were truly awful products. Possibly the worst cpu offered since the original no-cache celerons, and a terrible decision considering AMD's ability to offer low-cost Athlons, durons, and later semprons that absolutely whipped the celeron. If there was an engineering failure of the P4, it was complete inability to strip cache (saving a lot of manufacturing cost) and still maintain usable performance.

I am going to agree with the part about the NetBurst Celerons based on the 20 Stage design with only 128KB of LV2 cache. Those were indeed very slow in comparison to what AMD offered. Celeron D's are actually decent performers.

 

[LittleNemoNES]

lol sickbeast you must be restless -- making all these topics
:: Points to topic @ video forum ::

 

[SickBeast]

Originally posted by: Idontcare
I liked my P4, it did the job for me. No complaints.

As to whether it was an engineering failure...not sure what profession you harken to, but I am an engineer and there is nothing about the P4 that failed.

It did exactly as engineered to do. If it failed from engineering then it would not function.

Maybe you don't understand what engineering is or what failure is?

Your thread title is akin to stating mathematics is a failure because 2+2 = 4 when what you personally wanted was 1+1 = 2. Yes 2+2=4 and yes 1+1=2, neither is a failure and neither is bad mathematics. But if you personally wanted "2" then don't add 2+2 to get 4 and then claim mathematics is a failure.

The P4 is exactly as it was engineered to be. Whether you wanted a P4 or not is probably the only question you should ask yourself.

(No I do noyt and have not ever worked for Intel, but I am an engineer in the semiconductor industry)

Engineers are so rigid and take things way too literally.

By 'failure', I meant that it failed to live up to expectations. It did not give intel the performance crown for the majority of its production years.

I realize that true 'engineering failure' is a bridge collapsing or something.

 

[Fox5]

Originally posted by: Agentbolt
Not a diamond member, but it's hard to argue against AMD increasing their marketshare severalfractionsduring the P4 era. Wouldn't have happened if the P4 had been able to run faster for the price.

Fixed.

I voted yes, but only in the yes that it failed to meet its design goals. Add some water cooling in and the design had legs. Wasn't really a failure until the Prescott follow up anyway.

It definitely was overly ambitious and that came back to bite intel in the bum several times.

 

[SickBeast]

Originally posted by: gersson
lol sickbeast you must be restless -- making all these topics
:: Points to topic @ video forum ::

lol

We need *something* to do in the video forum! The place is dead!

It's not like I can create another R600 thread. :Q

 

[3chordcharlie]

Originally posted by: coldpower27

Originally posted by: 3chordcharlie
It was an engineering failure in terms of not doing what intel set out for it to do.

It was a product failure in terms of being beaten by Athlons for a large part of its run. But it was never beaten badly, and at certain times was the best product available.

The P4 turned out to be a poor design, but it wasn't an overall failure - it was a reasonable product.

I would suggest that netburst-based celerons did more to lose intel market share than P4 did. They were truly awful products. Possibly the worst cpu offered since the original no-cache celerons, and a terrible decision considering AMD's ability to offer low-cost Athlons, durons, and later semprons that absolutely whipped the celeron. If there was an engineering failure of the P4, it was complete inability to strip cache (saving a lot of manufacturing cost) and still maintain usable performance.

I am going to agree with the part about the NetBurst Celerons based on the 20 Stage design with only 128KB of LV2 cache. Those were indeed very slow in comparison to what AMD offered. Celeron D's are actually decent performers.

They did indeed, but by then the damage was done. Even now it's hard to argue with a $30-40 chip that runs in the same motherboards as high-end dual core parts.

 

[LittleNemoNES]

Originally posted by: SickBeast

Originally posted by: gersson
lol sickbeast you must be restless -- making all these topics
:: Points to topic @ video forum ::

lol

We need *something* to do in the video forum! The place is dead!

It's not like I can create another R600 thread. :Q

:laugh:

 

[StopSign]

It wasn't a failure. From a technical perspective it was a crappy chip but it did what Intel wanted: boost clock speeds. Back then all that people cared about in a chip was the clock speed, and most people nowadays still do. It gave people what they wanted and it was a chip that sold very well despite its technical shortcomings.

Intel lost market share because AMD came out of nowhere with the K8, just like how Intel came out of nowhere with Conroe.

 

[secretanchitman]

im using my p4 2.4C @ 3.6Ghz right now. its fast enough for what i need/want it to do.

northwood "C" p4s > any pentium 4 i think.

 

[Regs]

I don't think it was a engineering failure. I think marketing and production did most of the dirty work in taking down whatever hope Intel had with the P4.

In fact - the engineers had to redesign the chip 7 times so that it would be more cost effective and easier to produce for better profits. We all remember one of these redesigns called the williemate.

After that marketing took over and thought it would sell a lot better if they kept adding more and more. This was one of their more realistic internal goals. Make it more feasible for the public to believe that by always adding more brings a better product. Their goal wasn't to reach the 10GHz mark, that's just ludicrous. Of course the public bought into it because Intel can afford to shove it down customers throats.

I think it was a combined failure, if you want to call it that . Intel was more worried about reaching that status quo. They now seem to have switched to a more dynamic company and things seem to be clicking much better. Not to mention that they no longer pertend that AMD does not exist anymore.

 

[myocardia]

Originally posted by: Fox5

Originally posted by: Agentbolt
Not a diamond member, but it's hard to argue against AMD increasing their marketshare almost exactly fivefoldduring the P4 era. Wouldn't have happened if the P4 had been able to run faster for the price.

Fixed.

Fixed that for you.

Originally posted by: secretanchitman
northwood "C" p4s > any pentium 4 i think.

No doubt about it, the Northwood C's were the pinnacle of the P4; everything after that was a bit inferior to it. Unfortunately for Intel, there quite a few years between the Northwood C's and the Conroes.

Originally posted by: Regs
Their goal wasn't to reach the 10GHz mark, that's just ludicrous. Of course the public bought into it because Intel can afford to shove it down customers throats.

Whether they were lying or not, it's what Intel said. Well, at least it's what I remember them saying at one point. It seems like it was right before the introduction of the Preshotts.

 

[coldpower27]

Originally posted by: myocardia

Originally posted by: Fox5

Originally posted by: Agentbolt
Not a diamond member, but it's hard to argue against AMD increasing their marketshare almost exactly fivefoldduring the P4 era. Wouldn't have happened if the P4 had been able to run faster for the price.

Fixed.

Fixed that for you.

Originally posted by: secretanchitman
northwood "C" p4s > any pentium 4 i think.

No doubt about it, the Northwood C's were the pinnacle of the P4; everything after that was a bit inferior to it. Unfortunately for Intel, there quite a few years between the Northwood C's and the Conroes.

Originally posted by: Regs
Their goal wasn't to reach the 10GHz mark, that's just ludicrous. Of course the public bought into it because Intel can afford to shove it down customers throats.

Whether they were lying or not, it's what Intel said. Well, at least it's what I remember them saying at one point. It seems like it was right before the introduction of the Preshotts.

Let's see:

AMD marketshare Q1 2001 = 20.80%
AMD marketshare Q3 2002 = 11.60%
AMD marketshare Q4 2006 = 25.30%

Total X86 Marketshare numbers.

The actual correct statement would be.

Originally posted by: Agentbolt
Not a diamond member, but it's hard to argue against AMD increasing their marketshare over twofold during the P4 era. Wouldn't have happened if the P4 had been able to run faster for the price.

Five fold is flat out incorrect, nowhere in the past 7 years did AMD have only 5% marketshare. Revisionist history playing aren't we?

Cedar Mill's/Presler's on the 65nm process were decent performers and superior to Northwood's overall.

 

[secretanchitman]

Originally posted by: myocardia

Originally posted by: Fox5

Originally posted by: Agentbolt
Not a diamond member, but it's hard to argue against AMD increasing their marketshare almost exactly fivefoldduring the P4 era. Wouldn't have happened if the P4 had been able to run faster for the price.

Fixed.

Fixed that for you.

Originally posted by: secretanchitman
northwood "C" p4s > any pentium 4 i think.

No doubt about it, the Northwood C's were the pinnacle of the P4; everything after that was a bit inferior to it. Unfortunately for Intel, there quite a few years between the Northwood C's and the Conroes.

Originally posted by: Regs
Their goal wasn't to reach the 10GHz mark, that's just ludicrous. Of course the public bought into it because Intel can afford to shove it down customers throats.

Whether they were lying or not, it's what Intel said. Well, at least it's what I remember them saying at one point. It seems like it was right before the introduction of the Preshotts.

personally, when i heard that prescott was supposed to offer increased performance...i was ready to upgrade to that, until i saw how power-hungry and inefficient it was compared to athlon64s. still, i believed (still believe) in my northwood "C" p4. intel just increased clockspeeds and didnt even care about the other factors. of course, that is where the p4 came an "engineering failure", but i wouldnt call it that. i would call it...laziness on intels part. the cedar mill prescotts were actually decent cpus, as well as the presler pentium d's, but they should have been what intel released in the first place, not the prescotts and the smithfields.

man, did smithfields suck.

 

[judasmachine]

Originally posted by: secretanchitman
im using my p4 2.4C @ 3.6Ghz right now. its fast enough for what i need/want it to do.

northwood "C" p4s > any pentium 4 i think.

yeah the woody Cs were good chips, but IIRC they were rather expensive compared to amd's chips of the time. i got rid of my tbred 1700+ for a 3.0C, and never looked back. well i wish the p4 would have oced as well as the 1700+ but it was still faster so... (at a helluva price differance though.)

 

[Cookie Monster]

Northwood wasnt cosidered tobe better than prescotts and its other derivatives because it was "faster" in performance. It was "better" because it had a very good performance/power consumption ratio compared to the prescott. They also produced less heat, so it was a better overall complete package. If you look closely at old benches, prescotts exceeded the northwood in overall performance (e.g 3.6ghz perscott would be faster than 3.6ghz northwood although the difference isnt breath taking).

 

[myocardia]

Originally posted by: coldpower27
Let's see:

AMD marketshare Q1 2001 = 20.80%
AMD marketshare Q3 2002 = 11.60%
AMD marketshare Q4 2006 = 25.30%

Total X86 Marketshare numbers.

The actual correct statement would be.

Originally posted by: Agentbolt
Not a diamond member, but it's hard to argue against AMD increasing their marketshare over twofold during the P4 era. Wouldn't have happened if the P4 had been able to run faster for the price.

Five fold is flat out incorrect, nowhere in the past 7 years did AMD have only 5% marketshare. Revisionist history playing aren't we?

Cedar Mill's/Presler's on the 65nm process were decent performers and superior to Northwood's overall.

Got a link for that? I find it extremely hard to believe that AMD's marketshare has risen only 4.5% since Q1 2001. I mean, at that time, there were no companies (at least companies of any size, selling any/many systems) putting them into pre-built systems, and they definitely didn't have a large portion of the server market. So, I loved to see a link, proving that I'm the one doing the revisionist history.

And for anyone interested in what someone who gets paid to discuss processors and computer components has to say about the P4, this is a pretty interesting read.

 

[Aluvus]

Originally posted by: Cookie Monster
Northwood wasnt cosidered tobe better than prescotts and its other derivatives because it was "faster" in performance. It was "better" because it had a very good performance/power consumption ratio compared to the prescott. They also produced less heat, so it was a better overall complete package. If you look closely at old benches, prescotts exceeded the northwood in overall performance (e.g 3.6ghz perscott would be faster than 3.6ghz northwood although the difference isnt breath taking).

Below about 3.2 GHz, Northwood Cs typically outperformed Prescotts by a bit. At higher clock speeds, Prescotts took the lead. Similar to the difference between late-model Pentium IIIs and Pentium 4 Willamettes.

 

[coldpower27]

Originally posted by: myocardia

Originally posted by: coldpower27
Let's see:

AMD marketshare Q1 2001 = 20.80%
AMD marketshare Q3 2002 = 11.60%
AMD marketshare Q4 2006 = 25.30%

Total X86 Marketshare numbers.

The actual correct statement would be.

Originally posted by: Agentbolt
Not a diamond member, but it's hard to argue against AMD increasing their marketshare over twofold during the P4 era. Wouldn't have happened if the P4 had been able to run faster for the price.

Five fold is flat out incorrect, nowhere in the past 7 years did AMD have only 5% marketshare. Revisionist history playing aren't we?

Cedar Mill's/Presler's on the 65nm process were decent performers and superior to Northwood's overall.

Got a link for that? I find it extremely hard to believe that AMD's marketshare has risen only 4.5% since Q1 2001. I mean, at that time, there were no companies (at least companies of any size, selling any/many systems) putting them into pre-built systems, and they definitely didn't have a large portion of the server market. So, I loved to see a link, proving that I'm the one doing the revisionist history.

And for anyone interested in what someone who gets paid to discuss processors and computer components has to say about the P4, this is a pretty interesting read.

Here's a link, there is nothing to believe about it.

http://news.com.com/2100-1001-269940.html

AMD's market share actually had fallen since Q2 2001 where they were at 22.2% down to a low of 11.6% in Q3 2002, and over that time has since risen to 25.3% so at best even considering the time when AMD was at it's worse point for marketshare the increase is only a bit over two fold.

Since I have actually sourced my information now. I want to see a direct link on who has actually said that they made the P4 was designed to scaled to 10GHZ, without a link regardless of it is true or not cannot be used in this thread.

 

[StopSign]

Originally posted by: Regs
In fact - the engineers had to redesign the chip 7 times so that it would be more cost effective and easier to produce for better profits. We all remember one of these redesigns called the williemate.

Wasn't Williamette the original P4?

 

[yh125d]

It was Intel's push for clock speeds that makes people who dont know much about computers hesitant to buy a new 2.4gHz e6600 to replace their old 3.06gHz p4

 

[RichUK]

This is about the 50th reincarnation of this type thread.

Do we really need to keep going over the same stuff? Is there seriously anything more we can debate over?