Athlon 2700+ outperforms P4 3.06 HT

Copenhagen

Distinguished
Oct 21, 2001
552
0
18,980
I'm surprised to see that the combination:

<b>Athlon XP2700+ / ASUS A7N8X Deluxe (nForce 2)</b>

outperforms

<b>3.06GHz HT / ASUS P4PE (845PE)</b>

in all the benchmarks shown in Toms new VGA charts. Both systems were equipped with 512 MB RAM (DDR 333 - 2/2/2/5).

AMD beats Intel with:

<b>3.2 %</b> in Aquanox
<b>8.1 %</b> in Dungeon Siege
<b>2.5 %</b> in UT 2003 Antalus
<b>0.1 %</b> in 3D Mark2001 SE
<b>1.7 %</b> in Jedi Knight II 1.03

What's going on ? Frank Völkel's and Co. recent article showed that a P4 3.06GHz HT on an ASUS P4T533-C (Intel 850E chipset) and some PC1066 32ns RDRAM could beat an Athlon XP2700+ on an A7N8X and 2 x 256 MB DDR 400, Corsair, CL 2.0, PC 3200. In 3DMark 2001 SE Intel was 3.2 % faster.

This goes to show that a PC1066 RDRAM solution is still the ONLY combination to secure a victory over AMD.

With Intel blowing their chance to really dominate AMD with the Granite Bay, I certainly hope that Santa will be dumping an ASUS A7N8X Deluxe paired with an unlocked AMD Athlon 2400+ down the chimney.



<i>/Copenhagen - Clockspeed will make the difference... in the end</i> :cool:
 

eden

Champion
Yes I was surprised but then recalled it was a DDR setup. Still, the PE should've performed better than this IMO.
Nevertheless, it's even worse when you think that PC1066 is very expensive and just not worth throwing all the many in, so in the end you need DDR setups to be able to get a decent Intel system. In the end you got a 2.4 OCed to 3GHZ, behaving nearly like a PC1066 system, while the XP2700+ cost you probably around the same. If you get an XP2400+ and OC it, you might even get the chance to equal that 3GHZ OCed system.

That article at least showed what the majority of people use, and it seemed to me that AMD still does rule overall in performance-wise if the majority of people can't afford PC1066. This is also taking in account the XP2700 is not very expensive in contrast to the 2.8GHZ from Intel, and still performs better under DDR nForce 2 setups.
It was a new one to me though heh.

--
<A HREF="http://forumz.tomshardware.com/modules.php?name=Forums&file=faq&notfound=1&code=1" target="_new">The THGC Photo Album, send your pics and see others'!</A>
 

FUGGER

Distinguished
Dec 31, 2007
2,490
0
19,780
I am positive the results would be different on a IT7 Max 2 rev 2 as the P4PE is a dog in comparison.

The P4PE has a known problem with the ram dividers being incorrect. and there is a huge loss in memory bandwidth than what is advertised.

I have been working extensively with both boards to know exactly what I am talking about.

As my example I will use 3Dmark since it shows the desparity clearly.

<A HREF="http://service.futuremark.com/compare?2k1=5182699" target="_new">P4PE at 3.8Ghz scores 13477</A>

<A HREF="http://service.futuremark.com/compare?2k1=4548787" target="_new">IT7 Max at 3.37Ghz score 13227</A>

Video card clocks are almost identical, same DDR ram timing 2-6-2-2 yet the IT7 with almost disadvantaged by 500Mhz still competes head on with the P4PE. The P4PE is ok, but the IT7 and many other boards wipe the floor with it due to the problem with memory bandwidth.

I blame the loss in tests to the P4PE and not the 3.06HT

You are limited to what your mind can perceive.
 

FUGGER

Distinguished
Dec 31, 2007
2,490
0
19,780
My P4PE could not break my current high score @ default made by the IT7 with over 500Mhz advantage.

The above score were done at high res to tax the system as much as possible. I also ran FSAA benchmarks that had the exact same outcome.



You are limited to what your mind can perceive.
 

Copenhagen

Distinguished
Oct 21, 2001
552
0
18,980
I am positive the results would be different on a IT7 Max 2 rev 2 as the P4PE is a dog in comparison.
In all the reviews I've read, the Asus P4PE does well. I've never heard about this issue before. Could you try to explain in detail what the problem is ?


<i>/Copenhagen - Clockspeed will make the difference... in the end</i> :cool:
 

eden

Champion
I was about to say so as well, the P4PE did so well in all reviews, I don't understand where his performance drop is seen.

--
<A HREF="http://forumz.tomshardware.com/modules.php?name=Forums&file=faq&notfound=1&code=1" target="_new">The THGC Photo Album, send your pics and see others'!</A>
 

FUGGER

Distinguished
Dec 31, 2007
2,490
0
19,780
As you know I test a lot of products, and currently own a IT7 and a P4PE.

The P4PE has s problem with the memory dividers being incorrect or there is a loss of bandwidth available. In sandra you get a nice number but when running an application this number does not add up.

As my benchmarks show, my video card clocks are almost identical when you look at "fills" yet at nearly the same FSB the P4PE needs 500Mhz to compete with the IT7. If I ran both at the same speed in Mhz the P4PE is a joke. The advertised memory bandwidth is not there. Im using Corsair PC3500C2 and Geil 3500C2. I can get the P4PE up to speed no problem and it runs everything fine except apps that are bandwidth intensive. That is where it lacks.

<A HREF="http://www.xtremesystems.org/forums/showthread.php?s=&threadid=5613&highlight=P4PE" target="_new">Another P4PE thread</A> where JCViggen (current #1 in the world) also confirms P4PE has problems. I can find you more testimonials on the P4PE if you want.

You are limited to what your mind can perceive.