• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Toms Review: P4 3000 vs. XP 2300+

Page 2 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.


<< The P4 needs 1100 more mhz to overtake the Athlon. Funny >>



Mind you that there is a huge bottleneck. We just have to wait for the technology to overcome it.

-DocSmarts
 
Docsmarts... My PC2100 DDR is running at 156FSB which is near PC2700 Speeds. Thats 2.1gb/s vs 2.7gb/s theorotical bandwith. Sandra shows my memory scores at 2400/2400 WITH ddr at those speeds, much higher than the 2000/2000 shown for the 3ghz cpu here.Thats why using the asus p4b266 board,with that 3ghz would allow around 2600/2600 memory because their system bus on that is 137+33 which is 170FSB for the DDR. The option in the board allows the DDR to run at 133FSB while the CPU at 100. Then when you overclock the FSB the DDR goes with it. DDR is the bottleneck at PC2100(133FSB) not PC2700(166FSB).


Edit: The P4 didnt need 1100mhz it needed 466mhz. The 2.4ghz beat the athlon XP 2300+ in almost except the synthetic cpu benches and 1 or 2 application tests.
 


<< Docsmarts...did you read what I said....? My PC2100 DDR is running at 156FSB which is near PC2700 Speeds. Thats 2.1gb/s vs 2.7gb/s theorotical bandwith. Sandra shows my memory scores at 2400/2400 WITH ddr at those speeds, much higher than the 2000/2000 shown for the 3ghz cpu here.Thats why using the asus p4b266 board,with that 3ghz would allow around 2600/2600 memory because their system bus on that is 137+33 which is 170FSB for the DDR. The option in the board allows the DDR to run at 133FSB while the CPU at 100. Then when you overclock the FSB the DDR goes with it. DDR is the bottleneck at PC2100(133FSB) not PC2700(166FSB). >>



Christoph-

I am sorry that I did not understand your most advanced writing convention of memory benchmarks. You don't have to go postal on me. 🙂

Yes, I agree. The fact that the memory is the bottleneck is what I said above.

By the way, does your ASUS board have PCI and AGP locks? If not, how do you over come them from being overclocked?

-DocSmarts
 


<< Christoph-

I am sorry that I did not understand your most advanced writing convention of memory benchmarks. You don't have to go postal on me. 🙂

Yes, I agree. The fact that the memory is the bottleneck is what I said above.

By the way, does your ASUS board have PCI and AGP locks? If not, how do you over come them from being overclocked?

-DocSmarts >>



I think he is just frustrated as am I that Tom has once again pulled a dirty trick to prove his AMD bias.

He choked the P4@3000 instead of really unleashing its potential while he held nothing back on the XP.

And saying he didn't know about it isn't an excuse with his level of knowledge.

 
Yes it does have a 1/4 PCI AGP divider but only at 133FSB. It doesnt seem to be much of a problem now though as im running 123FSB right now and can do 127 FSB but with a hit to my memory benchmarks.
 


<< Edit: The P4 didnt need 1100mhz it needed 466mhz. The 2.4ghz be >>



Dunno, in only test that has some kind of practical value to me, the linux kernel compilation test, P4 running ~1133Mhz faster won with 1 second in front of AthlonXP. Not very impressive, IMO.
 
On the note of RDRAM and DDR, dual channel RDRAM takes 1/2 the circuitry on the motherboard PCB than single channel DDR SDRAM. So it'd make more sense to use a quad channel RDRAM (quad parity). The P4's FSB can be increased to 200MHz quad-pumped and you get 6.4 GB/sec.
 


<<

<< The P4 needs 1100 more mhz to overtake the Athlon. Funny >>



Mind you that there is a huge bottleneck. We just have to wait for the technology to overcome it.

-DocSmarts >>




Sorry. That is no excuse. We are compairing now, not in the future. You can say all you want to about the future, but you don't know what it will bring. We have now so we compare now. Also you can say all you want about the so-called bottleneck. The P4 has plenty of bandwidth. There are not many real world benifits as of right now.
 


<< So, we can take the numbers Tom gave us for the P4 at 2600 true but the P4@3000 is bugus because Tom did not overclock the Ram like MANY of the i845 motherboards can. >>

See here's the problem with overclocking the fsb in a comparision like this. You are correct about the 3:4 fsb/mem ratio, butI will point out that he probably on the 850 platform lowered the RDRAM ratio to 3x and thus the RDRAM was running at PC800. I really would've rather see equal fsb clocks/mem clocks, on the fastest platform for both CPU's, in other words KT266A running with PC2100 for the AXP and 850 with PC800 for the P4. But oh well, that's not allways possible. I do think that a better comparision needs to be made, but will it? prolly not. I will also point out though that despite Texmasters benchmarks, I find it unlikely that this is a conspiracy for Tom's AMD bias because he didn't even write the article. Frank V&ouml;lkel and Bert T&ouml;pelt did. So, I think I am just gonna dismiss these benchmarks
 
formulav8: I'll say this for the 3rd time...the P4 2.4ghz beat the XP 2300+ except in 3-4 tests. It doesnt need 1100 mhz.

Athlon4all: Huh? He had the RDRAM FSB at 120x4 =480 . Thats why he got 2900/2900 in the memory benchmarks. The other P4's were not overclocked much thus getting just above their normal 2400/2400 score.
 


<<

<<

<< The P4 needs 1100 more mhz to overtake the Athlon. Funny >>



Mind you that there is a huge bottleneck. We just have to wait for the technology to overcome it.

-DocSmarts >>




Sorry. That is no excuse. We are compairing now, not in the future. You can say all you want to about the future, but you don't know what it will bring. We have now so we compare now. Also you can say all you want about the so-called bottleneck. The P4 has plenty of bandwidth. There are not many real world benifits as of right now. >>



Did you not read the benchmarks? The 2600 overtook the 3000 many times.

The only difference was non overclcoked DDR ram vs Rambus.

Tom did not overclock the DDR ram like you can with most i845 chipsets.

His bias just rings true.
 


<<

<< So, we can take the numbers Tom gave us for the P4 at 2600 true but the P4@3000 is bugus because Tom did not overclock the Ram like MANY of the i845 motherboards can. >>

See here's the problem with overclocking the fsb in a comparision like this. You are correct about the 3:4 fsb/mem ratio, butI will point out that he probably on the 850 platform lowered the RDRAM ratio to 3x and thus the RDRAM was running at PC800. >>



I seriously doubt it because I got my P4 to run at 2.6 on my old i850 before converting to DDR.



<< I will also point out though that despite Texmasters benchmarks, I find it unlikely that this is a conspiracy for Tom's AMD bias because he didn't even write the article. Frank V&ouml;lkel and Bert T&ouml;pelt did. So, I think I am just gonna dismiss these benchmarks >>



I'm talking about the site not the person. Ever since anyone can remember, Tom's site has clearly had an AMD bias to it.

And the benchmarks I provided with 3dmark2001 show that overclocked DDR ram is extremely close to Rambus.

I believe the deceit was intentional.
 


<< formulav8: I'll say this for the 3rd time...the P4 2.4ghz beat the XP 2300+ except in 3-4 tests. It doesnt need 1100 mhz.

Athlon4all: Huh? He had the RDRAM FSB at 120x4 =480 . Thats why he got 2900/2900 in the memory benchmarks. The other P4's were not overclocked much thus getting just above their normal 2400/2400 score. >>




You can say it how ever many times you want and I could care less. I just read the 1100mhz from a post above and thought it was funny. The truth still is that the P4 needs a nice increase in mhz(rpm's) to beat the athlon even though is has double the L2 cache. So, say all you want. It still doesn't make the P4 look impressive. Especially for its overpriced price tag.



Jason
 


<<

<< formulav8: I'll say this for the 3rd time...the P4 2.4ghz beat the XP 2300+ except in 3-4 tests. It doesnt need 1100 mhz.

Athlon4all: Huh? He had the RDRAM FSB at 120x4 =480 . Thats why he got 2900/2900 in the memory benchmarks. The other P4's were not overclocked much thus getting just above their normal 2400/2400 score. >>




You can say it how ever many times you want and I could care less. I just read the 1100mhz from a post above and thought it was funny. The truth still is that the P4 needs a nice increase in mhz(rpm's) to beat the athlon even though is has double the L2 cache. So, say all you want. It still doesn't make the P4 look impressive. Especially for its overpriced price tag.



Jason >>



LOL boy talk about backtracking. Is it that hard to admit you were wrong about the Rambus vs the DDR?
 
Texmaster His bias just rings true.

Nah.

On one side I agree, seems they did not put as much effort into tweaking the P4 DDR platform - that is if it's true the Asus Board you mention can drive the memory bus async so easily - and without loosing much speed compared to synchronous operation. I have no experience of that board...

BUT, as jpprod already mentioned, the entire benchmark suite is very much geared for the P4.

Newtek's rendering app, for example, is not that much quicker on P4 because it's SSE2 optimised, but because on non Intel chips it apparently runs ancient MMX code (if there's any optmisation at all).
Of all the MPEG2 encoders available, they choose Pinnacle's, a company partly owned by Intel. What about Sysmark? Is that SSE patched for the Athlon? Why the selection of Light04 from the Viewperf suite?

I don't think one can claim an AMD bias here (or it's a change of fancy and they stick to their benchmark suite for 'consistency' while aiming for the results they whish to get?).
 


<<

<<

<< formulav8: I'll say this for the 3rd time...the P4 2.4ghz beat the XP 2300+ except in 3-4 tests. It doesnt need 1100 mhz.

Athlon4all: Huh? He had the RDRAM FSB at 120x4 =480 . Thats why he got 2900/2900 in the memory benchmarks. The other P4's were not overclocked much thus getting just above their normal 2400/2400 score. >>




You can say it how ever many times you want and I could care less. I just read the 1100mhz from a post above and thought it was funny. The truth still is that the P4 needs a nice increase in mhz(rpm's) to beat the athlon even though is has double the L2 cache. So, say all you want. It still doesn't make the P4 look impressive. Especially for its overpriced price tag.



Jason >>



LOL boy talk about backtracking. Is it that hard to admit you were wrong about the Rambus vs the DDR? >>





Excuse me? When did I say anything about Rambus vs DDR? You must be thinking of the wrong person.
 


<< Excuse me? When did I say anything about Rambus vs DDR? You must be thinking of the wrong person. >>



No I'm talking to you. Your comment was:

Also you can say all you want about the so-called bottleneck. The P4 has plenty of bandwidth. There are not many real world benifits as of right now.

And Its been proven that the memory bottleneck is very important and not overclocking the DDR to make up for that difference is a serious bottleneck.
 


<< Texmaster His bias just rings true.

Nah.

On one side I agree, seems they did not put as much effort into tweaking the P4 DDR platform - that is if it's true the Asus Board you mention can drive the memory bus async so easily - and without loosing much speed compared to synchronous operation. I have no experience of that board...

BUT, as jpprod already mentioned, the entire benchmark suite is very much geared for the P4.

Newtek's rendering app, for example, is not that much quicker on P4 because it's SSE2 optimised, but because on non Intel chips it apparently runs ancient MMX code (if there's any optmisation at all).
Of all the MPEG2 encoders available, they choose Pinnacle's, a company partly owned by Intel. What about Sysmark? Is that SSE patched for the Athlon? Why the selection of Light04 from the Viewperf suite?

I don't think one can claim an AMD bias here (or it's a change of fancy and they stick to their benchmark suite for 'consistency' while aiming for the results they whish to get?). >>



I'm going on his site's past dealings and this massive oversight about the DDR ram not being overclocked.
 
Formulav8: You said ...


The P4 needs 1100 more mhz to overtake the Athlon. Funny

I was merely correcting you saying that it needs 533mhz to overtake and from those benchmarks maybe even less. 2.4ghz and above the P4 beats the XP 2300+ . And duh it needs more mhz to compete,we've known that forever, but in no means does it need 1100mhz. It's not that funny.
 
BTW, I wasted a little time with Excel and keyed in the results, applying my own bias WRT benchmark selection: Lame, Linux, WinAce and DivX.

Slowest is the Athlon with 100 points, second the 3000A/DDR combo with 102.27, and the 2600A/RDRAM is speed king with 104.19.


 


<< The P4 needs 1100 more mhz to overtake the Athlon. Funny >>

and how much does it take to get a zealot's feathers ruffled? 😉

seriously though, i thought amd zealots preached that mhz isn't important? but now it is important to point out the mhz difference? or is it just important depending on which way your view is skewed?

intel makes the fastest cpu now. period. the best value for the money? nah, but definitely the fastest.
 
IMO, Intel and AMD people should agree on one thing, Toms hardware's reviews are becoming pathetic, Glaring mistakes, and a propensity to overstate things makes the time vested reading his articles = time WASTED.
Just my opionion.
please dont flame to much
 
Sid, I was replying to what a person said about the P4 needs 1100mhz more to beat the Athlon. So get it straight before you say anything.
About the mhz. Its the P4 owners praising the mhz. Like if they didn't think so much of the mhz why did they get all excited about a P4 overclocked to 2.8 or whatever it was for? Sounds like the rice boys that thinks their cars are fast because they can rev to 7k rpm. I could care less about the mhz of the P4 because the mhz of the p4 compared to a even clocked Athlon or P3 is pathetic and makes the P4 look aweful. So, I understand why no one cares about mhz anymore(Epsecially the P4 owners). The only time I look at mhz vs mhz is with the Athlon and P3. Intel definitely cares about the mhz or they wouldn't have developed the P4 core the way they did. Wonder why they would have done that for? Could it be because of the 99% of unknowing consumers? Naa. Intel would do anything like that.


Jason
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top