P35, X38, P45, X48 Crossfire Comparison

it's quite odd though as it shows P45 with 2 x8 PCI-E 2.0 outperforming X38 with 2 x16 PCI-E 2.0 slots so methinks something else is at play in their test.
 
im rebuilding my system in august, has anyone else got a link to a similar test ? want to see these results matching up on other reviews 1st before I pick my mobo
 
BUFF said:
it's quite odd ...
Click to expand...

To be honest it doesn't surprise me, PCIE 2.0 must have more improvements over 1.1 than just doubling the bandwidth, so these improvements could well make the 2.0 setup better than the equivalent bandwidth 1.1 setup.

Banjo
 
Banjo said:
To be honest it doesn't surprise me, PCIE 2.0 must have more improvements over 1.1 than just doubling the bandwidth, so these improvements could well make the 2.0 setup better than the equivalent bandwidth 1.1 setup.

Banjo
Click to expand...

...but P45 and X38 are are both PCI-E 2.0....

Here are compiled results:

World in Conflict:
WiC_Table.png


UT3:
UT3_Table.png


Supreme Commander:
SupCom_Table.png


Devil May Cry 4:
DMC4_Table.png


Crysis:
Crysis_Table.png
Click to expand...
 
Last edited:
What REALLY surprises me about those results is the P45 chipset actually showing a higher increase in performance compared to the X38!

Given that the X48 is the same chipset (essentially), why is the result always a LOT higher with the X48 compared to the X38?

Can someone please explain those to me please?

EDIT:
BUFF said:
it's quite odd though as it shows P45 with 2 x8 PCI-E 2.0 outperforming X38 with 2 x16 PCI-E 2.0 slots so methinks something else is at play in their test.
Click to expand...

Didn't read that first sorry buff... but am asking the question as to why :)

Matthew
 
I'm going to have to agree - something is not right there.
The X38 & X48 chipsets are the same - all that happened was that with X48 we got official 1600 FSB support.

I cannot believe the half-speed lanes on the P45 can out-perform the full-speed lanes on X38.
 
Update 16/07/2008: Previously we had found the performance of the Intel X38 chipset to be quite poor, which was surprising as the X48 really only brings 1600MHz FSB support to the table. However after spending hours upon hours testing with the ASUS Maximus Extreme, we were unable to beat the P45 performance with this X38 board. Therefore we ordered the Gigabyte X38T-DQ6 to re-test the X38 chipset, and as it turns out the results that we were receiving on the Maximus Extreme were much lower, as the X38T-DQ6 was able to match our X48 board.

So far we have been unable to work out why the Maximus Extreme is delivering such poor Crossfire performance and had initially suspected it to be a Catalyst driver issue, though we are now doubtful that the drivers are to blame. For those that had already read the article previously we do apologize. The Gigabyte X38T-DQ6 has been thoroughly tested over the past two days and we are now certain that we have true X38 Crossfire performance with the Radeon HD 4850 graphics cards.
Click to expand...

Explains it, but as it mentions.. very odd with the Asus board. BIOS issue?

Matthew
 
Last edited:
You must log in or register to reply here.
Back
Top Bottom