SLI Performance Previewed on X58 with GTX260 216SP and Forceware 180.32

[alexp999]

Expreview has got its hands on the NVIDIA Forceware Beta Driver 180.32 and used it to setup and enable SLI with two Galaxy GTX260-216, on an Asus P6T Deluxe motherboard. The cards scored a 3DMark Vantage score of P21623, compared against a single card score of P10920 gives an approximate 98% increase. Currently it is only possible to enable SLI officially on NVIDIA based chipsets, only AMD's equivalent Crossfire system works on Intel chipsets. It is evident that the 180 series of Forceware drivers from NVIDIA (dubbed Big Bang II) not only brings multi-monitor SLI support, but also paves the way to enabling SLI on the Intel X58 Chipset. In another article, Expreview have also mentioned that the beta driver 180.42, will be officially released from NVIDIA later on today, with the final version set for November 17th this year, coinciding with the launch of Intel's Core i7 and X58.



View at TechPowerUp Main Site

 

[phanbuey]

so basically they get 500 points more than my 192sp GTX260 SLI on 750i with less gpu score. Intel needs to make a dual core wolfdale with 12-16mb of cache at 4.5+Ghz for gamers... THAT would rock in games

 

[truehighroller1]

I get 100 points more with my setup. Q9450 @ 3.85Ghz, P45 Board, 4870 @ 840 / 1050, 4GB DDR2 1153Mhz~, 3 150GB drives in Raid 0.

 

[iStink]

I was hoping to see a jump in performance like conroe showed us. 3.2ghz at 1.18v is pretty wild though.

 

[phanbuey]

On the other hand, it is so cool to see Xfire and SLI on one board and it ACTUALLY WORKS... not like the skullripoff boards that intel released earlier. I bet this is the BIG BANG 2.

Although I wonder why they didn't use tri-SLI.

@ truehighroller1 - there is no way you get 100 points more in vantage with a single 4870.

 

[newtekie1]

I get 100 points more with my setup. Q9450 @ 3.85Ghz, P45 Board, 4870 @ 840 / 1050, 4GB DDR2 1153Mhz~, 3 150GB drives in Raid 0.

Proof.

 

[truehighroller1]

I meant then their single score of course and here is the link, also I was off a little they scored higher then my setup by 78 points and that is with me using the drivers without the physics support in other words if I use the physics drivers I would score higher then the single card setup here....


http://service.futuremark.com/compare?3dmv=451015

 

[alexp999]

I meant then their single score of course and here is the link, also I was off a little they scored higher then my setup by 78 points and that is with me using the drivers without the physics support in other words if I use the physics drivers I would score higher then the single card setup here....


http://service.futuremark.com/compare?3dmv=451015

This article is more about the SLI performance though. The single score is only given to show that the configuration of two cards has scaled properly.

 

[newtekie1]

I meant then their single score of course and here is the link, also I was off a little they scored higher then my setup by 78 points and that is with me using the drivers without the physics support in other words if I use the physics drivers I would score higher then the single card setup here....


http://service.futuremark.com/compare?3dmv=451015

We don't know if they had PhysX enabled or not. Plus, your processor and graphics card is overclocked, they ran everything at stock.

 

[truehighroller1]

I was close I'm sorry.. Still I score pretty high, you gotta give me props for that..

 

[Winterwind]

We don't know if they had PhysX enabled or not.

dont know?? do you really think that 3.2GHz icore 7 scores 44.5k cpu score w/o physx enabled?

 

[truehighroller1]

True that ^^ . The drivers have it enabled by default so I figured it was enabled off the bat..

 

[alexp999]

dont know?? do you really think that 3.2GHz icore 7 scores 44.5k cpu score w/o physx enabled?

would be nice. but I get about 33k CPU score on my Q6600 @ 3.2ghz with a GTX260 and GPU physx, so I would say its a pretty sure bet it is enabled.

 

[InnocentCriminal]

We've got a Q9450 stock (2GB DDR2800 - sales are stupid) with a 512MB ASUS HD4870 and that scored P9028, I've set it to run another 2 tests to see what we get again, but yeah I'm not actually impressed with their single card score, however the SLi score is pretty damn good.

 

[phanbuey]

We've got a Q9450 stock (2GB DDR2800 - sales are stupid) with a 512MB ASUS HD4870 and that scored P9028, I've set it to run another 2 tests to see what we get again, but yeah I'm not actually impressed with their single card score, however the SLi score is pretty damn good.

The single card score isn't great, thats for sure. I think they disabled PhysX for the single card run and re-enabled it for the SLi to try and boost the SLi score.

 

[InnocentCriminal]

Yeah, that's make sense. All we need now is some 4870's in Crossfire to compare.

 

[truehighroller1]

The single card score isn't great, thats for sure. I think they disabled PhysX for the single card run and re-enabled it for the SLi to try and boost the SLi score.

You might be right but, it still isn't that good even if it was disabled on the single but, mine is overclocked blah blah...

 

[phanbuey]

Yeah, that's make sense. All we need now is some 4870's in Crossfire to compare.

in vantage, you would be dissapointed. Sli spanks a 2way 4870Xfire in vantage, even with PhysX turned off. Thats because vantage sucks though, not because the cards are slower.

http://www.xtremesystems.org/FORUMS/showthread.php?t=195981 <-- thats with the 4870X2 in at 850MHz!!! only P16865 with a QX @4.4Ghz and a 16K GPU score (and its one of the highest ive seen, usually 4870X2 gets ~14-~15K GPU).

 

[truehighroller1]

in vantage, you would be dissapointed. Sli spanks a 2way 4870Xfire in vantage, even with PhysX turned off. Thats because vantage sucks though, not because the cards are slower.

http://www.xtremesystems.org/FORUMS/showthread.php?t=195981 <-- thats with the 4870X2 in at 850MHz!!! only P16865 with 16K GPU score (and its one of the highest ive seen, usually 4870X2 gets ~14-~15K GPU).

I noticed this as well.. Gimmicks I say.

 

[WarEagleAU]

Impressive with the scaling. This is how Xfire and SLI are supposed to perform.

 

[newtekie1]

The single card score isn't great, thats for sure. I think they disabled PhysX for the single card run and re-enabled it for the SLi to try and boost the SLi score.

That is entirely possible also. Questions like this are why I wait for the hardware to reviewed after release before getting excited about anything.

 

[alexp999]

The single card score isn't great, thats for sure. I think they disabled PhysX for the single card run and re-enabled it for the SLi to try and boost the SLi score.

I think its on for both. I get P11225 with my setup, and my Gfx card is OC'd.

 

[lemonadesoda]

I think 98% scaling is very impressive. I wonder if real-world performance scales so well. If so, hats off to SLI and the new drivers.

 

[DarkMatter]

I wouldn't expect the same scaling on games. Anyway has anyone ever compared SLI scaling in Vantage with the 790i, for example? I'm just curious now , though I never cared before .

 

[Hayder_Master]

don't forget guys physics take a good part from this score, if it run without physics it will give a cheap score , we know future mark know this physics trick and try to release banchmark don't depend too much on physics