2.72Ghz Conroe, X1900CF @ 712/738 takes AM3 WR on air

[njkid32]

I am getting really tired of seeing FCG having to defend his thread so if it doesnt stop he is going to have this thread closed and wont be posting his results in the open forums.. So please lets keep it clean and stop the flaming. Thanks...

[don_vercetti]

so glad i just sold my X2 system. Can't wait to hop onto the Conroe bandawagon ASAP.

And tbh, it's not really a big issue whether it fits it all into the cache or has very fast cache or whatever...certainly nothing to get worked up about. I don't see anyone here denying that conroe is faster than lightning, after all.

[Carfax]

I doubt you have any clue how a CPU works if you think the size of the EXE matters in the least when it comes to L2 cache usage in this benchmark.

What are you talking about? You're just babbling..

I never said the size of the EXE matters when it comes to l2 cache useage..

I said, that if the EXE can fit into the L2 cache, then the program will execute at a faster rate than if the main memory or hardrive were involved..

[PL4YD34D]

i believe that while the info here is undoubtedly important and thanks to FCG for doin these tests and providing results (and on WR), we have to ask some questions before rushing to sell grannys teeth to buy intel stuff.

first, if FCG has been given/allowed to show results with this chip, wuoldnt you make sure it was the best ES you could put out. that thing could be cherry picked from a thousand test samples.

second and more importantly, there is something fishy going on with CF working so damn good on these DDR2 chipsets. am i right in thinking that Conroe does not boast an on-die mem controller like the AMD's? so why suddenly so much more efficient? but it does have awesome bandwidth thru DDR2? that wouldnt just make up for AMD ownage for so long tho.

so somewhere round the middle, this setup (bad axe + conroe) is making up for the weak chipset structure that is found on AMD boards - given that there is no need for them to be manufactured with any real FSB control and throughput(due to on-die mem controller). This 'stronger' chipset structure used by the conroe(all credit to intel) is owning the graphics relay procedures ie driver function and PCI-e control (SB?) and allowing better bandwidth and functionality for the CF cards. well, says I anyways...

if this theory is correct (and please feel free to tear it to pieces), then AM2 may not show the better performance seen by the conroe camp because (i think??) the AMD's are keeping there on-die mem controller thru to skt AM2 boards which i believe will use 940pin layouts similar to the current Opteron server line. thus the chipset will not have made up the lost ground that could be slowing the whole CF system down.

Paul

[VulgarHandle]

amd started the ddr2 memory controller from scratch, and the 940 pin count is PURE coincidence

other than that, fun read

[ORCBEAST]

CONGRATS FCG Thx for your hard work and the World Record!
I appreciate it..As you know I was on this forum until 03:30 lol

Heres a BEER hehe.

Hopefully we can see some more benchies.
And I think that 2.4Ghz will be a real seller. It should go beyond 3Ghz easily. Dont you guyz agree ?

This chip FCG has will also go easily over 3Ghz with Air.Just the board that needs improvement, 2.7ghz is only the beginning ^___^


Regards
ORCBEAST

[PL4YD34D]

2) Memory bandwidth has nothing to do with CrossFire performance...at least not in the way you're thinking (it has an effect the same way it does on single card....CF doesn't utilize RAM much more than single card). The ATi chipsets are better than the Intel chipsets at CF. .

well you cant seriously tell me that i could run slower DDR2 against faster DDR2 and expect the same results - and then compare those to an AMD + DDR system of equivalent specification and get no difference again just because the cards were the same. i think the potential of the CPU to do work is somewhat reliant on its ability to use ram thus bandwidth affects the CPU bottleneck reduction of which you speak. and my point was not that the ram bandwidth has increased markedly, it was that the chipset had/ has changed sufficiently to affect the operation of CF

[Absolute_0]

It looks like it will take some people time to "accept" the results in their hearts

Anyway Kris the scores are absolutely great and i appreciate the work and the sharing. Can't wait for you to get a mobo that doesn't limit the OC.

[CPLB]

CONGRATS FCG Thx for your hard work and the World Record!
...

I agree, thanks a lot for putting this together FCG!

[mr_knowitall15]

wow i cant wait for these to come out, or at least mobos to support them so i could gat a 9xx to hold me over. I hope with a real OC friendly mobo you can get that "4Ghz easyily" that was promised by your guy.

[RimRam]

FCG, I apologies for sounding critical. Didn’t realize this thread was about WR. The FX60 screenshot threw me off. Again I Congrat you for your outstanding achievement. I know how you feel (I've been in the same situation), Sometimes its hard for some people to show a little appreciation.
the "Regal Crown" bestowed

[Pillo-kun]

for gaming i really dont really care so much about the cpu, i would rather pick a good video card instead. I justa wanna play without lag but for testing/beating records this cpu rox.

[R101]

Great work FCG!
Just unbelievable how Conroe breaks the WR without even getting a sweat.. Amazing!

[Donnie27]

Go ahead and ask anyone you know with the right cards to set their x1900 cards to 712 core and 738 mem and see if they can beat me. The cards are not what made this score possible. It was the CPU.

[Crush's voice from finding Nemo=Donnie27]Dewd yoooou RAWK![/Crush's voice from finding Nemo]

Congrats and keep up the very good work and your time spent doing it.

[taurboy]

The Intel is again the king, in the world of the computer, I hope that this new processors arrive soon to Spain. About the topic of the motherboard, it is know, If asus took out their soon his motherboard, with supports for the conroe?

[jakedeez]

OMG, thats all I have got to say. Rock on FCG, Intel and Conroe. I am gonna get maself some intel stock now.

I agree - INTC is a strong buy right now.

[integral]

FCG

Thanks for your great work. We all follow it closely here and look forward to more results.

Do not pay attention to all the fixed minds that find it hard to face the new reality. They have a problem and they are trying to make it yours. Just ignore them.

[railer]

FCG

Thanks for your great work. We all follow it closely here and look forward to more results.

Do not pay attention to all the fixed minds that find it hard to face the new reality. They have a problem and they are trying to make it yours. Just ignore them.

LOL

Really FCC you are tha man. You the first one to bring any meaningful results to us. Sure we'd like to see more. Do some CPu rightmark and some more 3dmarks. And don't forget some games if you can. Fear, q4, Farcry would be nice. Great job man and keep it coming

[zbogorgon]

FCG, one of your fanboys is convinced that u run CF with Dual 1900XTX without master card, so if I coudl asc you for a little favour. Please don't tell him the truth, it could have catastrophic consequences.

[NapalmV5]

LOL.. no matter how u dice it, slice it.. conroe is more powerful..

im no intel fanboy, neither amd, like both.. better 2 than 1 ..... cant believe all the negative talk about..

[rozzyroz]

hey FCG, have you tried this: ITK?

i really enjoy seeing these benchmarks bro, dont let all the haters bring ya down.

[Durzel]

Does all this talk of cache and SSE2 and what fits into Conroes cache and what doesn't really matter at the end of the day if it achieves the end goal of being the fastest in everything it does?

I'm all for synthetic benchmarks to draw comparisons, but to be honest Anandtechs article alone sold the Conroe to me - a min of 20% faster than an FX-62 in every single game they tried. 20%+ faster at a 1/4th of the cost - that's the bottom line as far as I'm concerned, everything else is a bonus. And frankly its obvious AMD have nothing to counter it this year, and what they decide to bring out next year is a whole different unrelated issue.

Obviously you can argue the legitimacy of Anandtechs review, and Intels showcase etc etc but these synthetic benchmarks that are coming out now are showing that they probably weren't faked, it really is going to be at least 20% faster than an equivalent setup with an FX-62.

Right now it seems we're just trying to find out just how much faster it can be than the existing stuff. It's not a question of "is it faster", its a question of "how much faster".

[VulgarHandle]

i want to know how amd can justify their prices with what yonah is doing, much less what conroe is going to do

[freecableguy]

FCG, one of your fanboys is convinced that u run CF with Dual 1900XTX without master card, so if I coudl asc you for a little favour. Please don't tell him the truth, it could have catastrophic consequences.

yep, x1900 series in CF requires a master card.

[DilTech]

i want to know how amd can justify their prices with what yonah is doing, much less what conroe is going to do

Same way intel justified their prices during the P4 vs A64 days?... It was the same, if not worse, than this upcoming one.

Either way... AMAZING FCG!