Phil_Boggild

But his Crysis benchmarks were run at 1920 x 1200... If he got 39 FPS at that res and you get 43 FPS at 1600 x 1200, it sounds like they're probably performing exactly the same as each other.

Anyway, totally pointless contributions aside, thanks to Rico for another benchmark. I think we all thought the 4870X2 would suffer more than the 4850X2. Crysis really must be the only game that uses enough vram to show any performance drop.

Very glad I've got 2 4890's on the way.

TraPofMinD

I must say that I am extremely interested in this thread! I have a Gigabyte EP45-UD3P and a 4890. I want to get a second one, but have been concerned about the split X16 issue.

I've seen lots of great simulations in this thread, but has anyone by now acquired two 4890's for crossfire tests on a P45?

I am still on the fence, but if I do purchase a second 4890, I will post my results.

rico2001

This thread already proves the 4890 doe not have any decrease in performance using 8X so you can expect the same results running two 4890s at 8X. Phil_Boggild had two of them but both were bad so no test from his setup.

__________________
My reviews:
4870, 4870 Tri-fire, 4850 X2 Quad-fire, 4850 X2 vs. 4870 X2,
4850 X2 + 4870 Tri-fire, 4850 X2 + 4870 X2 Quad-fire, 4850 + 4870 Crossfire,
(NEW) 4890 vs. 4850 X2

Steam nick: rico2001

"The ability to learn faster than your competitors may be only sustainable competitive advantage." ~Arie de Geus

d3v0

Rico, it still said 16x in gpu-z when you say it's 8x. Explain?

__________________

d3v0 - 16 Graphics Cards - Old/New Stock/Overclocked/Vmodded: Extensive Benchmarks

Grado SR-80s LNIB only $70 shipped!

CF cables, adapters, iPhone case, all sorts of stuff for sale CHEAP!!

rico2001

Well on Phil's post here:
http://www.overclock.net/6874421-post33.html
His 4870 is showing the gpu says it is on a pci-e 2.0 16x slot running at 8x.


On my 4870 post show here:
http://www.overclock.net/6879949-post43.html
I am also showing the card is on a pci-e 2.0 x16 slot running at 8x

Almighty15 did the same test with his 4890
http://www.overclock.net/6883656-post65.html
Same gpuz showing,pci-e 2.0 16x slot running at 8x.

Now for the X2 card, I believe gpuz show each gpu at 8x but since on the same slot, on same pcb, it can't tell them apart, added the x's together(8x/8x) and showed 16x. That is what I showed you all what my ccc showed, which was the X2 card at 8X.
http://www.overclock.net/6890198-post71.html
http://www.overclock.net/6899205-post78.html

__________________
My reviews:
4870, 4870 Tri-fire, 4850 X2 Quad-fire, 4850 X2 vs. 4870 X2,
4850 X2 + 4870 Tri-fire, 4850 X2 + 4870 X2 Quad-fire, 4850 + 4870 Crossfire,
(NEW) 4890 vs. 4850 X2

Steam nick: rico2001

"The ability to learn faster than your competitors may be only sustainable competitive advantage." ~Arie de Geus

d3v0

Awesome.

__________________

d3v0 - 16 Graphics Cards - Old/New Stock/Overclocked/Vmodded: Extensive Benchmarks

Grado SR-80s LNIB only $70 shipped!

CF cables, adapters, iPhone case, all sorts of stuff for sale CHEAP!!

Phil_Boggild

Just to clarify, the RMA of my 4890's had nothing to do with my P45 board. I tested them in another system under various conditions. A certain batch of XFX 4890's seem to underperform.

The P45 platform, as has been proven in this thread, is more than adequate for Crossfire .

Threefeet

Just want to say thanks for the thread guys, very interesting reading

+rep

__________________

3DMark06 : 23548

Screenshots

SuperPI Mod 1.5 - 1m : 10.920s