crossfire edition cards have been out since the X1900 days. The X1900GT, and all the 2900's etc only need the bridge, just like SLI, so u dont need a special card and can pair any too together. Makes much more sense!
Someday, they're going to have to stop making cards longer... or they're going to have to start making bigger cases. Seriously though, quad CrossFire X looks like absolute kickass. Now, if only they had some decent high-end cards to do it with.
I read somewhere that it actually has a six and an eight pin connecter just like the 2900xt. No way just one connector will power that thing. I have a full tower case and I think that it is even too long for it.
Just get a hack-saw to the case, not that hard! i no some people will moan but if your a case-modder should add a little fun cant wait to see benchys
Size of a brain as well = come to your senses by looking at the tests of 8800 vs 3870. ATi is slow and inefficient.uke2:
i think ATI did a smart thing.. what is more: Rich PPL that can afford 8800GTx's or ppl that only have $200 or so to spend..
LOL! Ahahaha, nice one mate.. So off-topic but hey! But yeah, looks impressive.. Gah that reply would work for that too ^^ "That's what she said".. Rofl.
Quad HD 3870 X2 ... just wow :banana:uke2: This thing would own Crysis. Probably 120FPS @ any resolution with AAx6 and AFx16 on ultra setting.
care to show us a link?I would assume that if it requires alot less power then the 2900 series I would think that one six pin connector can power it
Not so fast /* not gonna delete my following post but you should ignore it as it's wrong */ Ceteris Paribus , here comes some math. The 3850 was getting 85% of the performance of a GT, and getting 170%* over a single 3850s card with two of them in xFire. That means compared to a 8800GT, at 4 of them would be knocking down 340% *.85 = 289% versus a single 8800GT. The GT scales at 1.87 for sli so 289/187 implies 155%, so four of these cards in xfire will run 55% faster than 2 8800GT's. This is the high end of the prediction, the performance scaling of the cards is likely to deteriorate more with four cards than with two. http://www.anandtech.com/video/showdoc.aspx?i=3151&p=10 Using the same prediction technique 3 8800GT's should get ((1.87/2) * 3 ) = 281% 289/281 < 1.03 . ** Assuming you could get 3 way sli, 4 Xfire 3870's should be only slightly faster than 3 8800GT's in raw numbers less than 3%. Like I said Ceteris Paribus, all things equal, though naturally they won't be. *170% is the performance of two radeon cards in sli over a single card, * 2 to get 4 cards, *.85 to get performance compared to a 8800GT. ** (1.87/2) calculates the performance boost per card, * 3 for 3 cards. /my guess is 30-40% boost over a sli gt rig with QUAD 3870 unless a major driver change gives ATI a big boost. ETA: all that typing only to finally notice the article I based it on cited 3850's not 3870's
ok....:boozer: and what do you mean by "cited 3850's" ? I dont see anything about HD 3850s .They are dual HD3870 or am I reading something wrong?! :3eyes:
the sad thing is it probably still wouldn't be able to play it much over 16x10 with AA. the 3870 gets about 9 fps everything maxed at that resolution with no AA. lets say the 3870 X2 gets good scaling and goes up to 18 fps. Quad crossfire would then get about 2.5-3x that value so about 45-50 but that's without AA. Turn up the AA and it would get 35-40 avg tops. That's only at 16x10 with octo crossfire! Crysis is a beast! but a great looking one at that. a sexy beast like Austin Powers...sorry had to throw that in there