P3 Faster then P4???

[apoppin]

Originally posted by: Diable
A P3 is probably faster then a P4 of the same speed on old non-SSE2 enhanced code. But if the app has any SSE2 code the P4 will smack the P3 silly.

The Tualatins - even the Celerons - are enhanced over the regular Coppermine PIII and the PIII-M even more so.

 

[wetcat007]

What nonbenchmark app uses SSE2? And the P3 truely is significantly better in many ways, they would have been better off trying to keep updating that or something it was cooler, it was more power per the clock. THe P4 came out and basically denouncifed itself by the originals being slower than the P3 1GHz, well anyways this is one reason that AMD uses their clock speed system thing, becuz intel cores are now unefficent as can be.

 

[apoppin]

Originally posted by: wetcat007
What nonbenchmark app uses SSE2? And the P3 truely is significantly better in many ways, they would have been better off trying to keep updating that or something it was cooler, it was more power per the clock. THe P4 came out and basically denouncifed itself by the originals being slower than the P3 1GHz, well anyways this is one reason that AMD uses their clock speed system thing, becuz intel cores are now unefficent as can be.

You are forgetting that the PIII absolutely tops out well below the P4s 3.5Ghz (or so). And the P4 is designed to scale. Intel made a decision to "stick with" the P4 and and really hasn't looked back.

Although inportant, notebooks are not their main focus and it is very logical (and convenient) to devote some fabs to older but updated technology.

 

[wetcat007]

Originally posted by: apoppin

Originally posted by: wetcat007
What nonbenchmark app uses SSE2? And the P3 truely is significantly better in many ways, they would have been better off trying to keep updating that or something it was cooler, it was more power per the clock. THe P4 came out and basically denouncifed itself by the originals being slower than the P3 1GHz, well anyways this is one reason that AMD uses their clock speed system thing, becuz intel cores are now unefficent as can be.

You are forgetting that the PIII absolutely tops out well below the P4s 3.5Ghz (or so). And the P4 is designed to scale. Intel made a decision to "stick with" the P4 and and really hasn't looked back.

Although inportant, notebooks are not their main focus and it is very logical (and convenient) to devote some fabs to older but updated technology.

Yeah i know, but i was really trying to just say at the same clock speeds the P3 would be faster.

 

[apoppin]

Originally posted by: wetcat007

Originally posted by: apoppin

Originally posted by: wetcat007

Yeah i know, but i was really trying to just say at the same clock speeds the P3 would be faster.

You are stating the obvious. This has been well-known before the P-4 was released.

 

[imgod2u]

Originally posted by: wetcat007

Originally posted by: apoppin

Originally posted by: wetcat007
What nonbenchmark app uses SSE2? And the P3 truely is significantly better in many ways, they would have been better off trying to keep updating that or something it was cooler, it was more power per the clock. THe P4 came out and basically denouncifed itself by the originals being slower than the P3 1GHz, well anyways this is one reason that AMD uses their clock speed system thing, becuz intel cores are now unefficent as can be.

You are forgetting that the PIII absolutely tops out well below the P4s 3.5Ghz (or so). And the P4 is designed to scale. Intel made a decision to "stick with" the P4 and and really hasn't looked back.

Although inportant, notebooks are not their main focus and it is very logical (and convenient) to devote some fabs to older but updated technology.

Yeah i know, but i was really trying to just say at the same clock speeds the P3 would be faster.

First of all, how is that in any way significant? Secondly, as I pointed out, there's no neccessary reason the P4 should always remain slower on a clock-normalized scale than the P3. As software becomes more and more adjusted, we should see its clock-normalized performance increase. It already has quite a bit so far since 3 years ago (I pointed this out in a previous post). Give it another 3 years.

 

[bluemax]

I think it's safe to boil all this down to a simple conclusion:

Clock for clock, both chips measured at the same (ex. 1.4GHz), the P3 outperforms the P4.
One cannot simply say the P3 is better or P4 is better because the P4 has been released at clockspeeds ~2.5x higher than any P3.

For an absolute statement, you can safely say that the current P4 processors will greatly outperform the fastest P3.

Are we done now?

 

[dullard]

Originally posted by: bluemax
I think it's safe to boil all this down to a simple conclusion:

Clock for clock...the P3 outperforms the P4.

It really depends on the application. So lets not come up with one general conclusion like that. Tomshardware has a very good article comparing CPUs of recent times in several different types of programs. I used that article to do some clock-for-clock comparisons. Below are the clock-for-clock scores.

Here are programs where a higher number is better:

• 
  Program \ 3.06 GHz Northwood \ 2.0A GHz Northwood \ 2.0 GHz Williamette \ 1.3 GHz Williamette \ 1.4 GHz Tualatin \ 1.0 GHz Coppermine
  Quake 3 low res \ 119.1 \ 125.4 \ 114.6 \ 135.1 \ 131.1 \ 136.2
  Quake 3 high res \ 114.8 \ 124.1 \ 112.8 \ 133.5 \ 128.5 \ 132.8
  Unreal Tournament 2003 \ 69.4 \ 76.7 \ 69.3 \ 79.3 \ 93.7 \ 94.1
  3D Mark 2001 SE \ 5189 \ 6133 \ 5673 \ 6853 \ 7929 \ 8294
  PC Mark 2002 - CPU \ 2474 \ 2479 \ 2427 \ 2460 \ 3198 \ 2668
  SiSoft Sandra 2003 CPU - larger score \ 3134 \ 2516 \ 2470 \ 2469 \ 2751 \ 2729
  Sysmark 2002 overall \ 96.1 \ 105.0 \ 94.0 \ 106.9 \ 101.4 \ 97.0

Here are programs where a lower number is better:

• 
  Program \ 3.06 GHz Northwood \ 2.0A GHz Northwood \ 2.0 GHz Williamette \ 1.3 GHz Williamette \ 1.4 GHz Tualatin \ 1.0 GHz Coppermine
  WinRAR 3.0 \ 17.3 \ 38.5 \ 42.0 \ 81.5 \ 62.9 \ 109.0
  MP3 \ 23.5 \ 58.0 \ 59.5 \ 139.2 \ 115.7 \ 238.0
  MPEG-2 \ 95.6 \ 266.7 \ 273.5 \ 626.1 \ 435.2 \ 851.8

Lets put those numbers into words:

• 
  The P3 wins in these:
  In Quake 3, the early Williamette and P3 were tied and they beat the newer P4s clock-for-clock by roughly 10%.
  In Unreal, the P3s have a good 30% clock-for-clock lead over the P4s.
  In 3D Mark, the P3s have a good 40% clock-for-clock lead over the P4s.
  In PC Mark, the P3s lead by 10% to 30% clock-for-clock.
  
  They tie in these:
  In SiSoft, the lead varies: the top P4s lead by 10%, and the P3s lead by 10% over the bottom P4s.
  In Sysmark, I have to call it a tie.
  
  The P4 leads in these:
  In WinRAR, the P4 dominates the P3 clock for clock by up to 500%.
  With MP3s, the P4 wins by up to 900% clock-for-clock.
  With MPEG-2, the P4 wins by up to 800% clock-for-clock.

Conclusion. The P3 has a small but noticible clock-for-clock lead in some programs, they are tied in others, and the P4 just blows the P3 away in the rest. Of course this is all clock-for-clock. Mulitply by the higher frequency and the P4 will win in all tests.

 

[bluemax]

Works for me.

 

[fkloster]

Originally posted by: merlocka

Originally posted by: fkloster

I haven't really looked into it myself, but it's true EXCEPT for video/audio manipulation (A/V Editting mainly)....in those cases the P4 spanks all.

FACTS:

1) the HIGHEST clock P4 (3.06) FLOGS the HIGHEST clock P3 (1.0) by an EXTREME margin in every test...

2) the LOWEST clock P4 (1.3) FLOGS the LOWEST clock P3 (.45) by an EXTREME margin in every test...

...any questions?

Yes. I believe the question was "is the p3 faster than the p4 mhz for mhz".

Thanks for pointing out the obvious though. It was very helpful.

Funny, I was under the impression that the title to this thread was "P3 Faster then P4???" (misleading as that may be)

Bottomline? The author of this thread is just trolling the same old 'clock for clock' crap we used to argue with over AMD and Intel.

CLOCK FOR CLOCK comparisons are irrelevant and crap @ best...