Triple-GPU Scaling: AMD CrossFire Vs. Nvidia SLI

Name: Test Settings - Triple-GPU Scaling: AMD CrossFire Vs. Nvidia SLI
Author: Thomas Soderstrom

by Thomas Soderstrom March 14, 2011 at 12:00 AM

Test Settings

Test System Configuration
CPU	Intel Core i7-2600K (Sandy Bridge) LGA 1155: 3.40-3.80 GHz, 8 MB L3 Cache Overclocked to 4.00 GHz at 1.25 V
Motherboard	Asus P8P67 WS Revolution: LGA 1155, BIOS 1007 (01/24/2011) Intel P67 Express, Nvidia NF200 PCIe Bridge (2 x PCIe 2.0 x16)
RAM	Kingston KHX2133C9D3T1K2/4GX: 2 x 2 GB, DDR3-2133 at DDR3-1600 CAS 7-7-7-21, 1.60 V
Hard Drive	Western Digital WD1002FBYS: 1 TB, 7200 RPM, SATA 3Gb/s, 32 MB cache
GeForce Graphics	Asus ENGTX570/2DI/1280MD5: 742 MHz GF110 GPU, 1280 MB GDDR5-3800
Radeon Graphics	XFX HD-695A-CNFC: 800 MHz Cayman GPU, 2 GB GDDR5-5000
Sound	Integrated HD Audio
Network	Integrated Gigabit Networking
Power	OCZ-Z1000M: 1000 W Modular, ATX12V v2.2, EPS12V, 80 PLUS Gold
Software
OS	Microsoft Windows 7 Ultimate 64-bit
GeForce Graphics	GeForce/Ion 266.58
Radeon Graphics	AMD Catalyst 11.1
Chipset	Intel INF 9.2.0.1019

CPU “bottlenecks” are always a problem at medium (and lower) resolutions. Today’s tests start at medium resolutions, so we clocked Intel’s Core i7-2600K to 4.0 GHz in an effort to minimize that issue.

The Achilles-heel of Intel’s fastest processor is a platform that supports only sixteen PCIe 2.0 lanes directly from the processor to the graphics card. Asus solves that solution by incorporating Nvidia’s 32-lane NF200 bridge, thereby repeating the same information at full bandwidth to two cards simultaneously. Triple-GPU mode operates in an x8/x8/x16 configuration, as it would with most X58 motherboards.

Power might be the price for performance, but we wanted to see if it also scaled with performance as GPU numbers increased. OCZ’s 80 PLUS Gold-rated Z1000M provides that clean power while adding only around 11% overhead at our watt meter.

Benchmark Configuration
3D Games
Aliens Vs. Predator Benchmark	Alien vs Predator Benchmark Tool Test Set 1: Highest Settings, No AA Test Set 2: Highest Settings, 4x AA
Crysis	Patch 1.2.1, DirectX 10, 64-bit executable, benchmark tool Test Set 1: Highest Quality, No AA Test Set 2: Highest Quality, 4x AA
F1 2010	v1.01, Run with -benchmark example_benchmark.xml Test Set 1: Highest Settings, No AA Test Set 2: Highest Settings, 4x AA
Just Cause 2	Version 1.0.0.2, Built-In Benchmark "Concrete Jungle" Test Set 1: Highest Details, No AA, 16x AF Test Set 2: Highest Details, 4x AA, 16x AF
S.T.A.L.K.E.R.: Call Of Pripyat	Call Of Pripyat Benchmark version, all options, HDAO Test Set 1: Ultra Preset, DX11 EFDL, Ultra SSAO, No AA Test Set 2: Ultra Preset, DX11 EFDL, Ultra SSAO, 4x MSAA
Synthetic Benchmarks and Settings
3DMark 11	Version: 1.0.1.0, Benchmark Only

Previous Next

Summary

About the author

Thomas Soderstrom @hardware_tom

Thomas Soderstrom is a Senior Staff Editor at Tom's Hardware US. He tests and reviews cases, cooling, memory and motherboards.

This thread is closed for comments

163 comments

Comment from the forums

Your comment

amk09

i wish i could afford that
tacoslave

wow even though a 6950 is weaker than a 570 it dominates at higher resolutions eyefinity here i come.
scrumworks

Impressive and unexpected results. Speculations that crossfire scaling is worse than SLI can finally be put into rest.
aznguy0028

scrumworksImpressive and unexpected results. Speculations that crossfire scaling is worse than SLI can finally be put into rest.

Not true, I would have liked to see more games on the test. The choices were too limited for my taste. For example, everyone knows that Crysis is heavily Nvidia favored, so it's not surprising to see those results in the least.

Here's another article on the 68xx series in Xfire
http://www.xbitlabs.com/articles/video/display/radeon-hd6870-hd6850-crossfirex.html#sect0

Xfire scaling was found to be 100% on 10 or the 19 games they tested, and remaining very high on the others. Nvidia and ATi seems to trade blows at different games tested, but it is evident that Ati has stepped up their Xfire drivers int he 6xxx series.
aznguy0028

scrumworksImpressive and unexpected results. Speculations that crossfire scaling is worse than SLI can finally be put into rest.

I misunderstood what you meant. But yes, xfire/sli scaling has improved alot this generation
liquidsnake718

Now im thinking if i should even spend on a now older 5850 to crossfire on my X58.
bavman

Its really cool to see amd stepping up their game. Multigpu scaling used to be only good on nvidia, but now amd is beating them. Though i feel that more games should be tested before a firm conclusion should be drawn.
_Pez_

I would get two HD 6870 for a mobile crossfirex gaming rig !
anubis44

No! Don't buy a 5000-series card to crossfire. The 6000 series cards are essentially all better at crossfire scaling than their predecessors. This performance advantage will likely only increase with additional driver revisions and better 6000-series optimization support than for older products.
Maziar

Great review ! It's good to see AMD did a lot of work on CF scaling.
classicaxe

I wish you Tom's would have done a 6970 trifire comparison instead. It's on average cheaper than the 570 on the Egg right now. I guess i get the memo, with showing that a cheaper solution is outperforming a more expensive option, but there would have been a good deal of performance increase across the board with the clock speed differences and opened shaders that is the 6970. Those differences come more alive in scaling especially with 3 cards.
Rizlla

You can see that AMD knows multicard setups are the future. Now i really want to get my 6950
avatar_raq

@Thomas:
You can test eyefinity resolutions on single monitor IF that monitor has 3 inputs. Can you give this a try please?
anubis44

Holy crap! After reading this review, I decided to look into the 6950 cards again.

If you're in Canada, you can get an XFX 6950 1Gb card for $215.99 (after rebate)right now from NCIX! That's a stupidly amazing deal!!! Think I'll grab one!
http://www.ncix.com/products/index.php?sku=58455&vpn=HD695XZNFC&manufacture=XFX&promoid=1293
Luay

Loved the idea of the article. Very disappointed with the display setup. It should go up to 5760x1200.
Crashman

LuayLoved the idea of the article. Very disappointed with the display setup. It should go up to 5760x1200.

Asus' 25" 1920x1200 display is still well priced, but I don't expect to see them around much longer. You should tell Chris to order three of these for every editor, before that bargain price vanishes!
liquidchild

why not test the 6970 and the 570? they are the same price point, the 6950 is one tier lower.
Sihastru

Quote:

Things get ugly for Nvidia at 2560x1600. While a 67% performance gain for two-way and 112% gain for three-way SLI certainly aren't shabby, AMD’s 92% two-way and 245% 3-way gains put it in the winner’s circle.

Well the graph for nVidia shows 212.11%, so a 112% gain is correct, but the graph for AMD shows 244.87%, so 245% gain is incorrect, that should be 145% gain. Still better then nVidia, but not that much better...
executor2

"The closest competitor is Nvidia’s GeForce GTX 570" Wrong !

The closest competitor is the Nvidia`s GTX 560Ti price vs price. At least in my country.
Crashman

executor2"The closest competitor is Nvidia’s GeForce GTX 570" Wrong ! The closest competitor is the Nvidia`s GTX 560Ti price vs price. At least in my country.

Could you please show me the GTX 560 Ti model that supports 3-way SLI? I'd love to test those too!
Bigmac80

don't forget that 6950 2gb can be flashed to 6970
blibba

"A third card appears to be about as appealing as an auxiliary nipple at 1680x1050"
executor2

@Crashman

I`m sorry my mistake , I haven't looked up to the 560 if they are capable of 3-way crossfire, ALSO isn't the author`s job to make a balance article ?

Pitting two cards against each other should also have the same price ? This seems hidden commercial for Nvidia , just my 2 cent.
Crashman

executor2@CrashmanI`m sorry my mistake , I haven't looked up to the 560 if they are capable of 3-way crossfire, ALSO isn't the author`s job to make a balance article ? Pitting two cards against each other should also have the same price ? This seems hidden commercial for Nvidia , just my 2 cent.

Did you read the last three pages, where the cheaper cards win? I think most people would regard that as an ad for AMD! Or are we not reading the same article?

Now if AMD really wanted to bribe someone, some 6970's might have resulted in fewer complaints...but that's water under the bridge now.