[MaadDaawg]

That is the question.. whether tis greater among men to use the Classified with the NF200 chip, or faster using the Classified without the NF200 chipe? Having one of each, one which is hydro copper (non NF200) and one without a complete chipset water block (with the NF200) varily I asketh which board wouldst perform best upon the phased platform? As the one which shall be encassed in a steel HAF will demand full water cooling, methinks the best bet may be to use the power of bits to submerge the NF200 equipped Classy in the cooling waters as well? But yea I say unto thee, my research has not produced a clear winner of the speed crown betweenth the two and thus I am gazing upon yet another expenditure which I wouldst be better advised not to make.

LOL So, anyone read which board is faster under phase.. I haven't found anything that says yea or nea :(

[Chuchnit]

Why not test them up on a table? I have wondered the same thing, but people only make the claim the 760 is faster. It would be nice to see a "real world" experiment to see for sure. Maybe 03 and either vantage or 06 to test?

[tanka12345]

Isn't NF200 only needed if you want more than 3 cards?

I think the Eclipse Plus features NF200, but Classifieds are probably (or surely) more stable.

[MaadDaawg]

The NF200 is what allows you to run true 16x16x16 TRI SLI. With the other Classified boards, including the Hydro Copper you can run 16x in SLI only and 16x8x8 in TRI SLI. Rumor always was the NF200 added latency but I've never seen proof either way. I'm sending in my Ltd. Ed. (759) Classy for RMA and plan to water it when it comes back, then I'll have both versions with the chipsets watered, but, one will be under phase and the other under water so I won't be able to do a true match up :( Unless of course, someone wants to loan me their phase unit for the test lol

[Afterburner]

Other than maxing out my Quad 295's, I never have found the need for 16x... So today, it really does not matter...

We all know how each release adds one or two goodies... And the 5000 ATI's and NVidia 300's are gossiped to need this extra room (Bandwidth). Although I do not see any concrete proof yet...

I run both... and for me it is the luck of the run.. One week I think the None 200 is king because my benches get better... Then the next I am out scoring those with the 200... so.. Meh...

[MaadDaawg]

What you say AB conforms to what I've heard - only need the 16x bandwidth for the 295s. Luckily, all the classifieds will run quad sli at full 16x on both slots

[punx223]

i have run back to back with the 759 and 760 with quad 295's and have saw NO difference in vantage.... and mind you the 760 is 16/8 even in dual sli mode.... so 8x does not touch the 295....... now for future proofing I guess you could say future cards may be hurt by 8x, but my question is.... how much will they be hurt? let alone by the time such a card comes around will the classy still be the TOP BOARD or will EVGA have come out with a newer TOP DOG board at that time in order to keep the masses happy.

[Chuchnit]

Originally Posted by MaadDaawg

What you say AB conforms to what I've heard - only need the 16x bandwidth for the 295s. Luckily, all the classifieds will run quad sli at full 16x on both slots

Originally Posted by punx223

i have run back to back with the 759 and 760 with quad 295's and have saw NO difference in vantage.... and mind you the 760 is 16/8 even in dual sli mode.... so 8x does not touch the 295....... now for future proofing I guess you could say future cards may be hurt by 8x, but my question is.... how much will they be hurt? let alone by the time such a card comes around will the classy still be the TOP BOARD or will EVGA have come out with a newer TOP DOG board at that time in order to keep the masses happy.

There ya go MaadDaawg. Plus you gotta think about this. Say a single 295 can almost 100% saturate pci-e 2.0 x16 slot. It would require 100% scaling in SLI/Xfire for it to saturate the second one. Plus for benchmarking that will never happen.

Another "theoretical" thing to think about the nf200 chip. Any dual card solution from nvidia will already have an nf200 on it for internal SLI. So in "theory", the nf200 board SHOULD be slower with 295/ATI dual card solution (PLX chip) because of not only the added latency of one nf200, but three if you run two dual gpu cards. Now personally I think the only time you'd notice is IF you cared about 3-7fps in a game you are over 125+ on already. I really don't think you would notice until you started playing high resolution games.

[sofos1990]

Whatever you do, NF200 is useless.
We are talking about "virtual 4x16 PCI-E"
All depend on X58:
1 vga = 1 x16 PCI-E
2 vgas = 2 x16 PCI-E
3 vgas = 2 x16 + 1 x4 PCI-E or 1 x16 + 2 x8
4 vgas = 4 x8 PCI-E
So how can NF200 handle 4 x16 PCI-E?
It's useless, it adds only latency.
All above are hipro5's words from a greek forum.....here

[punx223]

actually on the E760 classified

4 vga =
1x16
1x8
1x4
1x4


on the 760 even with 2 vga's you get 16x and 8x

And i have said it before and will say it again. I have run 759 and 760 back to back with quad 295 and saw no difference in vantage