AMD vs Intel dual cores?

[IntelUser2000]

Cinebench is REALLY not a good benchmark. Go see multi-processor scaling for Cinebench and you'll see most CPUs scale at the same rate, which is 84-86% for 1 to 2 CPUs/cores.

Saying that X2 is better than C2D because of Cinebench is really idiotic IMO. Comparing between two systems from forums is also BS because they aren't exactly identical setups. You can see in reviews Cinebench scores are slightly better for C2D than X2. That slight difference can be negated 2-3 times over by comparing systems that are not exactly identical(like comparing two systems from two different users in a forum).

All other benchmarks show 20%+ advantage per clock for C2D over X2. What more do you need? Rely on one flawed benchmark??

 

[Duvie]

Originally posted by: IntelUser2000
Cinebench is REALLY not a good benchmark. Go see multi-processor scaling for Cinebench and you'll see most CPUs scale at the same rate, which is 84-86% for 1 to 2 CPUs/cores.

Saying that X2 is better than C2D because of Cinebench is really idiotic IMO. Comparing between two systems from forums is also BS because they aren't exactly identical setups. You can see in reviews Cinebench scores are slightly better for C2D than X2. That slight difference can be negated 2-3 times over by comparing systems that are not exactly identical(like comparing two systems from two different users in a forum).

All other benchmarks show 20%+ advantage per clock for C2D over X2. What more do you need? Rely on one flawed benchmark??

I can attest to this cause my 4 cores scale so great in cinebench yet in real world apps of 3d rendering I cannot get anywhere near to full 3 core use let alone 4...This was tested on AMD dual opterons as well....nothing real world in rendering that I have tested scales like cinebench. YET!!!!


http://www23.tomshardware.com/cpu.html?modelx=33&model1=466&model2=432&chart=188

18.4% faster clock for clock for the C2D

http://www23.tomshardware.com/cpu.html?modelx=33&model1=466&model2=432&chart=177

32.9% faster clock for clock for the C2D

http://www23.tomshardware.com/cpu.html?modelx=33&model1=466&model2=432&chart=187

31.4% faster clock for clock for the C2D

http://www23.tomshardware.com/cpu.html?modelx=33&model1=466&model2=432&chart=176

33.8% faster clock for clock for the C2D


See a pattern?

http://www23.tomshardware.com/cpu.html?modelx=33&model1=466&model2=432&chart=175

old stable of AMD....17.3% faster clock for clock for the C2D

http://www23.tomshardware.com/cpu.html?modelx=33&model1=466&model2=432&chart=166

16% faster clock for clock for the C2D



http://www.techreport.com/reviews/2006q4/core2-qx6700/index.x?pg=4
http://www.techreport.com/reviews/2006q4/core2-qx6700/index.x?pg=5

15.3%, 26.6%, 17.0%, 9.8%, and 12.4% respectively in clock for clock advantage for C2D

http://www.techreport.com/reviews/2006q4/core2-qx6700/index.x?pg=8

2.3%, 22.8%, 14.1% and 15.0% respectively clock for clock advantage for C2D



Seems like a mass conspiracy....

 

[Roguestar]

Well duh, we all work for intel, remember?

 

[uo7]

Quick question. I'm looking to update my htpc. I just installed the xbox360 hd-dvd drive and it choked horribly. Amd mobile barton 2600. To keep things cheap is there a E6300 mobo out there that still takes older DDR ram ? and possibly agp. In my opteron it runs smooth as well. Just trying to decide on a cheap way to upgrade it for HD. I already tested it in a friends E6300 and even without hardware accel on it runs flawless.

 

[Roguestar]

Yes, there are a few motherboards that will support AGP and PCI-E, and DDR I and DDR II. I haven't read much about them personally but anandtech did a series of articles on the ASRock 775Dual-VSTA which you can find here:
http://www.anandtech.com/mb/showdoc.aspx?i=2814

 

[uo7]

Originally posted by: Roguestar
Yes, there are a few motherboards that will support AGP and PCI-E, and DDR I and DDR II. I haven't read much about them personally but anandtech did a series of articles on the ASRock 775Dual-VSTA which you can find here:
http://www.anandtech.com/mb/showdoc.aspx?i=2814

Thanks for the reply, that board looks perfect for my needs.

 

[lopri]

Originally posted by: Roguestar
What cooling are you using on that, Lopri?

I used to use Ultra-120 with P5W-DH. Now I use Scythe Infinity for EVGA 680i board. The capacitors on the back of the board was the main reason for the switch. At first I didn't like the mounting mechanism of the Scythe, but now I've become to like it and the cooling performance seems to be slightly better than Ultra-120.

 

[lopri]

I sort of disagree with the opinions above regarding Cinebench. Well, more like in general (or in theory, if you so prefer). My beef with the Cinebench wasn't that it was a bad benchmark, but it just doesn't bear much resemblance to an 'average' user's desktop usage. (just like Super Pi doesn't) I mean, who needs quad, octa cores for desktops? (those who actually do, please excuse me. But you guys sure know that you're minority! ) But for theoretical tests, I think it's a wonderful little app that can show the power of multi-cores.

If an application (which is meant to be work in a multi-core enviroment) doesn't scale as well as Cinebench, I'd rather blame the app instead of Cinebench. Don't get me wrong - I understand that the actual applications are billion times more complicated than Cinebench. But the truth that i believe is, the better scale the apps are, well, the better they are. For those folks who do utilize many cores, Cinebench can be a very useful benchmark to gauge the power of their systems. (and quality of the apps to some extent)