Booj

Hey guys, as some of you would know, a brand spanking new Kayl cascade and a pair of HD4890 video cards arrived this week. After a good bench session, i've got a few results to share

This is my first ever sub zero bench session hehe.. Insulation and setup was a breeze, thanks to the kneaded erasers and kayl's mounting and backplate. It got even easier when the Q820A599 E8600 i am using showed no cold bug problems at all. After about 4 or 5 hours of benching, everything was still alive! The cold bug switch for this cascade hasn't needed to be used. I did try it out though. It sounds awesome

Setup:



E8600 Q820A599
Asus P5E3 Premium bios 0704
2x1gb G.Skill 1800C7 D9JNM
2x Sapphire HD4890 reference cards w/ stock air coolers
WD 74gb Raptor older one.
Seasonic M12 700w

Best 1m didn't save it properly It was 7.6x seconds at 6050mhz

Max Valid 6133mhz


Pifast


The main course. These are all untweaked runs. I'm going to learn how to tweak 3d hehe
HD4890 Crossfire 3d Mark 2001


03


05 Really happy with this one


AM3


There is more to come. More single card and 3d06. I'll take my 4850's for a spin sometime and switch over to the i7 for 32m

Thanks for looking guys (Eva feel free to move to another section)

eva2000

Sweet results mate.. you're in the right section

So how cold did your first session go for evap/cpu ?

Booj

coool cheers

My thermometer is an el cheapo jaycar special. I'm not sure if its entirely accurate. It hovers about -98 to 96 most of the time and drops to around -92 at 3d mark 2006 cpu test.

chispy

Awsome results Booj , if you turn the switch for coldbug off it will go down how much ?
-50C more or less ?

Booj

Thanks man. We both did some hwbot damage today hehe

Kayl reported it dropped to around -80 but my thermometer says dropping to around -70. I'm really not sure how accurate mine is at these low temps. It is calibrated with ice water.. I don't wanna spend the $ on a fluke

scottath

Nice scores.
Why only 1 CF bridge? when i ran crossfire (HD3870's) 2 was faster - and it makes sence if you use 2 it would be faster (more data bandwidth??)

Any results of mixed crossfire yet? (4980 + 4870/4850) if the drivers will allow?

scottath

Booj

Thanks, I'll try that. I wasn't aware the 2 made any difference

hehe no 4850/4890 testing yet. I've still got a lot more to bench with this platform with 4850 crossfire too and then switch it all over to i7

scottath

i think it is more noticable the difference at higher resolutions - but still made ~300mrak difference at 1280*1024 in 06 (from memory)

When you get a chance.....i would so love to be able to afford such systems.....