yes my ran on slot nearest cpu, slot1 i think (4870x2 on slot1, 4870/512 on slot3(another x16 slot)) tryed to get it run tri on xp so i tested every combination but that was no go reason i ever bought that mobo was able to do 2x16 pci-e and maybe get all three slots filled, but as that really hard to find info about tri/quad fire and xp was just mumbojumbo from amd and theres no real enviroment effects on vista/win7 with x-fi... well u can see my rig on left (just gave win7 7057 a test run and stutter was horrible and yet no proper sounds)
OKay guys i figured out why there was so much blurring and stuff when i ran vantage using the x2 as the 1st card. It also happened when i changed the postiion of the cards. This is cos i own 2 OC cards but they are OCed at different speeds. Okay my 4870's mem is not OCed. my x2 is at 800 and my 4870 is at 770 and i dun really think my 4870 can OC till 800, well further testing needs to be done. ANyway here's my vantage scores again, they look fantastic don't they. my cpu score is pulling down the total though. tri-fire score: http://service.futuremark.com/compare?3dmv=957928 Just the 4870 x2 score: http://service.futuremark.com/compare?3dmv=949022
maybe running the X2 in primary slot does help. your GPU score should not pass mine when your CPU is clearly less effecient. unless the higher clock speed and bus speed gives the GPU's more headroom to talk to eachother. i dont own a 4870X2. i am thinking about buying one but they are still a lot of money here. as for you worries about weather a game supports crossfire or not. it doesnt really matter because the game will still only be using one GPU. the 4870X2 is a crossfire card on games that do not support crossfire at all the X2's used to actually score lower than the single GPU varient ... now this is becomming less and less likly since the drivers have been fixed and or games are supporting the tech. i am wondering how your trifire scored so much higher in test 1 and 2.. maybe if its not too much trouble could you clock down to 3550 ish for a clock to clock comparison ? its pretty flippin sweet IMO that graphically your trifire is faster than the one i set up. for my sake i hope its only drivers, fingers crossed.
Vantage doesn't scale gpu with cpu speed hence it is a more accurate measurement compared to 06 which scales the gpu according to the cpu speed and the type of cpu used. I was quite surprised that your tri fire and x2 score is quite low. What do u mean clock down to 3550? U mean downclock the x2 mem to 3550? The stock mem is at 3600, is urs at 3550?
your CPU is at 4000MHz isnt it? try clockign it down to 3550 ish to match my cpu and see what your GPU score is by comparison please.
Don't you have 4 cores?? my cpu score loses to urs when my clocks are at 4ghz. Vantage measures gpu speed seperately from cpu speed. i'll do it when i have the time.
sapphire 4870/512 and club3d 4870x2 all those gpus was good on 815 but not much further, finally had em on "safe" 800 memory on both cards was just misareble with 925 top oc, even auto oc from ccc gave more normal oc values but they just couldnt take it i consider both my cards was pretty bad clockers, better than some that cant oc at all tho
Bull. Nothing wrong with a 4870, or two, or three. Upgrading to a core i7 platform only means the scaling of ANY of the cards you have will work like they were intended to, but only failed on a 775 platform. That "cheapest parking space" you refer to is still at the forefront right now, and there is really no architectural difference with the 4890 either, so it eludes me as to what you are talking about. I have a core i7 with 2 4870s. And yes, I am in process of adding a third. and my "parking space" is right outside my office as far as I'm concerned. In closing, let's put it this way...To any corei7 platform owner, buying 4870s, or anything for that matter in xfire, or sli, and putting it on a 775 platform, is VERY much like your initial assertion only the opposite, except that in my version of your analogy, it holds true.
Have you tried cooling them wth anything other than air, and then upping the volts a tad? Something tells me you haven't.
THANK YOU. It's amazing how ATI releases a card with a higher nomenclature and everyone suddenly thinks a 4870gpu is ****. And a mass migration ensues. (through lga775 goggles) lol @ simple marketing and how it works! Even my 3870x2 cards actually worked great on the I7 and made very competitive numbers sometimes beating out some of the more modern stuff. Way better than they could dream of working on my q6600. On an i7 platform, yes it is worth it, because even what many here consider to be an "old" video card will work that much better, and thus negating buying the same damned card with a die shrink of a hefty premium. LOLZ Hilbert does dome damned hard work so you don't have to. They're pretty good reads. But of course they are not always the bottom line. They are general guidelines. I recommend people go out and touch the hardware themselves and experience it before conclusions are drawn. All that does is spread the disease among the less experienced.
if CPU MHz has nothing to do with it then why did your trifire outscore mine? http://service.futuremark.com/compare?3dmv=957928 http://service.futuremark.com/compare?3dmv=796033 so thats 12111 CPU and 17964 GPU for me and 8057 CPU 21503 GPU for you. where does that make any sense. the only thing different is my fsb was 200 where yours is around 444 ? therefor giving your trifire more bandwidth on the pci-e laneways.. its either that or its the drivers, or combination. i was/still am on 9.1 .. do the 9.4's seriously make a large improvement? if i could clock my cpu up to 4.0GHz i would. if i had an X2 i would also get the new drivers and retest. but i cant and i dont. thats why i asked you to downclock your CPU a bit, so i could see if its the architecture of core vs PII, or drivers, or somthing else.
Think it has more to do with the FSB overclock, indirectly the MHZ oc. His mobo is working faster then yours, together with higher overall clocks pushes his FPS high in vantage, if this is also true in a gaminging enviroment might not be true. Think best would be to compare Crysis, Farcry and world in conflict benches to compare fps to see wich system performs the best. Both Crysis and WiC is PCU driven games that love high clocks, where farcry is more gpu bound and (hopefully) will show a diffrence based on the pci-e lanes. But generally speaking I don't really know why his score is so much higher then yours, you are with 3fire at same levels as a single 295 GTX where his really outperforms it by a good margin.
yes i tested gpu volts up to 1.35(default max 4870x2 has in bios) temps was fine with tuned fan profile cant really test em further as i dont have em anymore
intel has more memory bandwidht and amd has better latency well.. i7 changes this as it has memory controller next to core like amd has had from athlon(?) also raw calculate power is approx double on intel yet amd is more towards integrers that should be better for gfx, cpu tests on 3dmarks leans towards raw calculate(this is something digged from darkest parts of my brains instead reliable sites)
Right, I finally found the test Ledhead linked to a few days ago. It's a review made by anandtech that compare Am/am2, Phenom X4, X3, C2D, C2Q and I7 in gaming enviroments as synthetic benches. Anandtech CPU review Enjoy! /edit Multi GPU test on C2D, C2Q and i7 clock for clock comparison /edit2, Really interessting result in the 2nd review if you link to all the tested games, C2D is really strong in clock by clock comparison vs all the quads and i7.
