P4@3700 against XP 2600+ and 2800+ on Sandra 2003

[Pedro Rocha]

I know Sandra benchmarks are not to accurate but here is how my P42800@3700 compare against AMD XP 2600+ and 2800+ on the new Sandra 2003 benchies .

AMD newcomers are not bad in CPU and Multimedia speed, but they really fall behind in Memory Speed - limitations on the VIA chipset..








Here is the ID and temperatures of my CPU:

[OPPAINTER]

Well here's my XP2400@2700,, 1000 less Mhz then your P4 and the CPU bench kicks your booty

OPP

[OPPAINTER]

Some MM bench at 2700Mhz

OPP

[OPPAINTER]

You got me on the mem bench though
XP2400 @ 2722 Mhz

OPP

[NoOne]

nice work OPP !

[OPPAINTER]

Thanks NoOne

Welcome to Xtreme

OPP

[QuadDamage]

OPP

these are some nice sandra benches! i need to get Prommy ASAP and at least try to get closer to you, my man

[NoOne]

Thnks OPP
Im Brasilian but i undestand english..
My only problem is "type in english" hehehehe

[DDTUNG]

Welcome to Xtreme, NoOne.

Look around our forums. There's a lot of useful information.

And feel free to join the Xtreme Folding@Home team.

DDTUNG

[petervandamned]

with pc800 instead of pc1066 is already fine

[QuadDamage]

hehe, nice too bad i can't find my sandra piccie of P4/DDR running at 243fsb well beating your rambus.

[Pedro Rocha]

Originally posted by OPPAINTER
Well here's my XP2400@2700,, 1000 less Mhz then your P4 and the CPU bench kicks your booty

OPP

Fantastic OC my friend and Guru..

But the numbers don't make much sense to me and really demonstrate the inaccuracy of Sandra Benchies..

Your XP2400@2700 gives 10.196MIPS and a PR3930, my P42800@3700 gives only 9.000 MIPS but a PR4077.

Definitively we need some more accurate benchmarks to compare both platforms..

[Raptor]

here's a score from my DDR system...

Using regular 256MB PC2700 Samsung D with relaxed timings...

http://pwp.netcabo.pt/raptor/sandra_176_234.jpg

And one from my old Rimms System...

http://pwp.netcabo.pt/raptor/sandramem.jpg


ps: had to put on links because the pictures are to big for 56k'ers...

[ronnie0738]

You guys need to compare by running spec benchmarks. :thumbsup:

[VEG]

Well here's my XP2400@2700,, 1000 less Mhz then your P4 and the CPU bench kicks your booty

I think i love you OPP :p

...just wish i had all the money you had, to upgrade every 15mins

[chainbolt]

Originally posted by ronnie0738
You guys need to compare by running spec benchmarks. :thumbsup:

or better application based benchmaks. Sandra is nice for a first look but not really showing what's going on, I mean nobody would seriously compare CPU power by running Sandra

Run Sysmark or Content Creations Winstone or other productivity
test and and you will see the difference between a XP2700+ and a P4 3700 Quite a substuntial one. I have compared a while a go a XP 2200+ with a P4 2.26 and they were basically at equal level, and now guess what will happen if you compare a 2700+ and a P4 3700 Mhz with applications. In most tests the XP 2700 will look very old. In Content Creation 2002 you will see at least a difference of 10%.

[]JR[]

Superpi/pifast and 3dmark are good comparisons, and with the former I dont think any p4 will keep up with a 2.8ghz amd, the latter can be seen on the orb, like 12th position beating everything upto 3.5ghz p4, thats likely to be mostly mem bandwidth limitations of amd compared to rambus or ddr pulled up p4 scores...

]JR[

[chainbolt]

Originally posted by ]JR[
Superpi/pifast and 3dmark are good comparisons, and with the former I dont think any p4 will keep up with a 2.8ghz amd, the latter can be seen on the orb, like 12th position beating everything upto 3.5ghz p4, thats likely to be mostly mem bandwidth limitations of amd compared to rambus or ddr pulled up p4 scores...

]JR[

We know since the launch of the K7 core that Athlons are more "productive", their ouput is higher at the same frequency. It's a stoneold story, and it does not help, becuase that is more than compensated by the much higher frequency the P4 core is capable of, and that's it. Therefore all top postions are hold by P4 systems and not by Athlons.

And the best is that we have a RDRAM system running at 138 MHz with a score above 20,000: that is still the direction where to look at. The top score is also hold by a RDRAM system. It's shame that this technology is put to rest now.

[NoOne]

Money is nothing if you don't have technique

[chainbolt]

Originally posted by NoOne
Money is nothing if you don't have technique

same goes for the girlfriends

[NoOne]

hehehe.. realy

[Jeff]

Originally posted by ]JR[
Superpi/pifast and 3dmark are good comparisons, and with the former I dont think any p4 will keep up with a 2.8ghz amd, the latter can be seen on the orb, like 12th position beating everything upto 3.5ghz p4, thats likely to be mostly mem bandwidth limitations of amd compared to rambus or ddr pulled up p4 scores...

]JR[

Or how about about SETI@home... P4's need over 1GHz to be competitive.

http://www.users.globalnet.co.uk/~mu.../benchmark.htm

I know that if you are purely into high performance then a P4 is hard to beat, but AMD is not too far behind and the CPU, mobo, and memory(not so much anymore, but RDRAM is still king) are all less $$$ so it's definitely a better bang for the buck with AMD.

[Jeff]

Originally posted by chainbolt
Run Sysmark or Content Creations Winstone or other productivity test and and you will see the difference between a XP2700+ and a P4 3700...

Since the P4 is overclocked, why would you compare it to a stock 2700+? That sounds like a marketing strategy.

If you want to be "fair", compare that P4 with a 2700+ that's overclocked to 2.8 or 2.9GHz... now that'd be something I'd like to see. I really don't think the P4 would show a substantial performance margin.

[Ewok]

it isnt 2700+ as they dont actually exist (or no doubt people would have them?) its 2.7ghz.

[GreenBeret]

From the P4 3.06GHz HT Sandra scores, I can see that P4 HT will laugh at AthlonXP in Sandra very soon.........