Multi-GPU SLI/CF Scaling: Lynnfield's Blemish

When running in single-GPU mode, the on-die PCIe controller maintains a full x16 connection to your graphics card:


Hooray.

In multi-GPU mode, the 16 lanes have to be split in two:

To support this the motherboard maker needs to put down ~$3 worth of PCIe switches:

Now SLI and Crossfire can work, although the motherboard maker also needs to pay NVIDIA a few dollars to legally make SLI work.

The question is do you give up any performance when going with Lynnfield's 2 x8 implementation vs. Bloomfield/X58's 2 x16 PCIe configuration? In short, at the high end, yes.

I looked at scaling in two games that scaled the best with multiple GPUs: Crysis Warhead and FarCry 2. I ran all settings at their max, resolution at 2560 x 1600 but with no AA.

I included two multi-GPU configurations. A pair of GeForce GTX 275s from EVGA for NVIDIA:


A coupla GPUs and a few cores can go a long way

And to really stress things, I looked at two Radeon HD 4870 X2s from Sapphire. Note that each card has two GPUs so this is actually a 4-GPU configuration, enough to really stress a PCIe x8 interface.

First, the dual-GPU results from NVIDIA.

NVIDIA GeForce GTX 275 Crysis Warhead (ambush) Crysis Warhead (avalanche) Crysis Warhead (frost) FarCry 2 Playback Demo Action
Intel Core i7 975 (X58) - 1GPU 20.8 fps 23.0 fps 21.4 fps 41.0 fps
Intel Core i7 870 (P55) 1GPU 20.8 fps 22.9 fps 21.5 fps 40.5 fps
Intel Core i7 975 (X58) - 2GPUs 38.4 fps 42.3 fps 38.0 fps 73.2 fps
Intel Core i7 870 (P55) 2GPUs 38.0 fps 41.9 fps 37.4 fps 65.9 fps

 

The important data is in the next table. What you're looking at here is the % speedup from one to two GPUs on X58 vs. P55. In theory, X58 should have higher percentages because each GPU gets 16 PCIe lanes while Lynnfield only provides 8 per GPU.

GTX 275 -> GTX 275 SLI Scaling Crysis Warhead (ambush) Crysis Warhead (avalanche) Crysis Warhead (frost) FarCry 2 Playback Demo Action
Intel Core i7 975 (X58) 84.6% 83.9% 77.6% 78.5%
Intel Core i7 870 (P55) 82.7% 83.0% 74.0% 62.7%

 

For the most part, the X58 platform was only a couple of percent better in scaling. That changes with the Far Cry 2 results where X58 manages to get 78% scaling while P55 only delivers 62%. It's clearly not the most common case, but it can happen. If you're going to be building a high-end dual-GPU setup, X58 is probably worth it.

Next, the quad-GPU results from AMD:

AMD Radeon HD 4870 X2 Crysis Warhead (ambush) Crysis Warhead (avalanche) Crysis Warhead (frost) FarCry 2 Playback Demo Action
Intel Core i7 975 (X58) - 2GPUs 25.8 fps 31.3 fps 27.0 fps 70.9 fps
Intel Core i7 870 (P55) 2GPUs 24.4 fps 31.1 fps 26.6 fps 71.4 fps
Intel Core i7 975 (X58) - 4GPUs 27.0 fps 57.4 fps 47.9 fps 117.9 fps
Intel Core i7 870 (P55) 4GPUs 24.2 fps 50.0 fps 36.5 fps 116 fps

 

Again, what we really care about is the scaling. Note how single GPU performance is identical between Bloomfield/Lynnfield, but multi-GPU performance is noticeably lower on Lynnfield. This isn't going to be good:

4870 X2 -> 4870 X2 CF Scaling Crysis Warhead (ambush) Crysis Warhead (avalanche) Crysis Warhead (frost) FarCry 2 Playback Demo Action
Intel Core i7 975 (X58) 4.7% 83.4% 77.4% 66.3%
Intel Core i7 870 (P55) -1.0% 60.8% 37.2% 62.5%

 

Ouch. Maybe Lynnfield is human after all. Almost across the board the quad-GPU results significantly favor X58. It makes sense given how data hungry these GPUs are. Again, the conclusion here is that for a high end multi-GPU setup you'll want to go with X58/Bloomfield.

A Quick Look at GPU Limited Gaming

With all of our CPU reviews we try to strike a balance between CPU and GPU limited game tests in order to show which CPU is truly faster at running game code. In fact all of our CPU tests are designed to figure out which CPUs are best at a number of tasks.

However, the vast majority of games today will be limited by whatever graphics card you have in your system. The performance differences we talked about a earlier will all but disappear in these scenarios. Allow me to present data from Crysis Warhead running at 2560 x 1600 with maximum quality settings:

NVIDIA GeForce GTX 275 Crysis Warhead (ambush) Crysis Warhead (avalanche) Crysis Warhead (frost)
Intel Core i7 975 20.8 fps 23.0 fps 21.4 fps
Intel Core i7 870 20.8 fps 22.9 fps 21.5 fps
AMD Phenom II X4 965 BE 20.9 fps 23.0 fps 21.5 fps

 

They're all the same. This shouldn't come as a surprise to anyone, it's always been the case. Any CPU near the high end, when faced with the same GPU bottleneck, will perform the same in game.

Now that doesn't mean you should ignore performance data and buy a slower CPU. You always want to purchase the best performing CPU you can at any given pricepoint. It'll ensure that regardless of the CPU/GPU balance in applications and games that you're always left with the best performance possible.

The Test

Motherboard: Intel DP55KG (Intel P55)
Intel DX58SO (Intel X58)
Intel DX48BT2 (Intel X48)
Gigabyte GA-MA790FXT-UD5P (790FX)
Chipset: Intel X48
Intel X58
Intel P55
AMD 790FX
Chipset Drivers: Intel 9.1.1.1015 (Intel)
AMD Catalyst 9.8
Hard Disk: Intel X25-M SSD (80GB)
Memory: Qimonda DDR3-1066 4 x 1GB (7-7-7-20)
Corsair DDR3-1333 4 x 1GB (7-7-7-20)
Patriot Viper DDR3-1333 2 x 2GB (7-7-7-20)
Video Card: eVGA GeForce GTX 280
Video Drivers: NVIDIA ForceWare 190.62 (Win764)
NVIDIA ForceWare 180.43 (Vista64)
NVIDIA ForceWare 178.24 (Vista32)
Desktop Resolution: 1920 x 1200
OS: Windows Vista Ultimate 32-bit (for SYSMark)
Windows Vista Ultimate 64-bit
Windows 7 64-bit

Turbo mode is enabled for the P55 and X58 platforms.

The Best Gaming CPU? SYSMark 2007 Performance
Comments Locked

343 Comments

View All Comments

  • nikrusty - Wednesday, November 18, 2009 - link

    With this article Anandtech is Harder, Better, Faster Stronger.
    Seriously AWESOME ARTICLE! It cleared many of my doubts FLAT OUT! Now I know i5 is the way to go especially becoz I dont care about overclocking and just want good gaming performance...nothing screamingly extreme. Budget + Performance always keeps you level headed.
  • shiro - Wednesday, October 21, 2009 - link

    what is that monster hoop of death heatsink that's on page 3? lol
  • Eeqmcsq - Saturday, September 19, 2009 - link

    I asked a similar question in one of the other articles, so pardon me if this sounds repetitive.

    According to the Turbo charts, the slowest Turbo speed is higher than the stock speed. Why is that? For example, why not just make the 750 a stock GHz of 2.8 GHz instead of 2.66GHz?
  • Eeqmcsq - Saturday, September 19, 2009 - link

    Argh, please ignore. Replied using the wrong Firefox tab.
  • The0ne - Tuesday, September 15, 2009 - link

    Clear up what you're trying to show on the graphs please. You're getting more FPS at max setting than at min settings? Label the graphs like you did with the others please. With the others I can just look and understand what you're doing. With these, I'm scratching my head.
  • The0ne - Tuesday, September 15, 2009 - link

    Ah, turbo mode represented in FPS >.>'
  • kkara4 - Monday, September 14, 2009 - link

    over at bittech.net, they are saying that it is more worth it to go for the i7-920, if we are considering anything above the i5. this is a conflicting story, since anand is recommending the lynnfields. anand or anybody else for that matter could you please see their articles and tell me what they have done wrong? (or perhaps you guys failed to see something). Your article explains things in great technical detail which i can understand since i have studied microprocessors, hence i am more inclined to go for lynnfield. anyway if someone could cross check that would be good
  • mapesdhs - Tuesday, September 15, 2009 - link


    If I've understood Anand's analysis correctly, the conclusion is that,
    for application mixes which involve a lot of single and/or dual-threaded
    codes, and assuming one is not interested in high-end SLI/CF setups
    or hard oc'ing all 4 cores all the time for tasks like video encoding
    or animation rendering, the 750/860 are better buys because they
    will internally push 1-core and 2-core clocks to a higher rate than
    occurs with the 920 via the Turbo function, giving better results
    than the 920, and of course the 750/860 are cheaper solutions
    (although the 860 price is similar to the 920, the mbd costs less
    than an X58, from what people say).

    So it depends on what you want to use your system for. No interest
    in CF/SLI? Running games that don't hammer 4 cores? An i5 750 or
    i7 860 makes more sense. Using apps that don't use more than 2 cores?
    Again the 750/860 is more logical, especially from a cost viewpoint.

    This ties in with the other advantage of the X58 platform, ie. the
    upgrade path to 6-core and 8-core CPUs. If this is something that
    holds no value to you, then P55 makes more sense.

    As always, it depends on what you want to use the system for. The
    attraction of the 860 from a more general point of view is that it
    also offers good quad-core performance when one does use all 4 cores
    without sacrificing the traditional higher-clocks possible with
    single or dual core setups when one is only using 1 or 2 cores. It's
    the best of both worlds, at least for out-of-the-box functionality
    anyway.

    However, if one does intend to use all 4 cores almost all the time
    (I do) with a strong overclock, then the 920 is a better choice
    because of the voltage issue and (IMO) the 6/8-core upgrade path.
    Likewise, high-end multi-GPU setups work better with X58.

    Given that general usage of a PC rarely uses more than 2 cores, this
    is why the 750 and 860 are such attractive options.

    As for the 870, despite its 1/2-core speed advantages, the price is
    too high IMO. For that kind of money, a 920 makes more sense, paired
    with better cooling if one has such a spare budget, or buy a better
    GPU setup which, for gaming, is where the real bottleneck lies.

    Anand, please correct me if I'm wrong with the above.

    Ian.

    PS. As always, real-world pricing issues can make a mess of on-paper
    technical conclusions. Also, although many games/apps don't exploit
    more than 2 cores now, this is likely to change in the near future as
    multi-core coding becomes more pervasive in the industry.

  • mapesdhs - Monday, September 14, 2009 - link


    Anand/Gary,

    Re your comments about an X58 advantage being the ability to use
    later 6 and 8-core CPUs...

    I've been planning to build an i7 920 system for video encoding, so
    a max oc on all cores is useful to me; from the article I thus infer
    the X58 is a better choice.

    However, if I did buy such a setup instead of an i5 or i7 860, what
    would the cost tradeoff be do you think when the 6-core CPUs arrive
    with respect to upgrading? By that I mean, for total processing
    throughput, do you reckon a 6-core upgrade would be significantly
    cheaper than simply buying a second i7 920 setup? (gfx not an issue)
    If not, then the ability to use 6/8-core CPUs later in this context
    is somewhat lessened, something that would apply to animation
    rendering aswell (ie. extra complete systems perhaps more cost
    effective in increased overall throughput compared to upgrading to
    more cores). Any ideas? Also, unless the applications used can
    exploit more than 4 cores, the later 6-core CPUs won't help. I have
    about 1500 hours of material to convert to DivX. Each file is about
    40 to 45 minutes (documentary), so converting multiple files on
    multiple systems at the same time is very doable.

    Given the above, I'm looking forward to more details on how a max
    oc'd i860/i870 compares to a max oc'd 920.

    At present I'm just using a 6000+ setup to work out the appropriate
    format/conversion paths.

    Ian.

    PS. May I suggest you don't bother replying to those moaning in such
    an obviously ludicrous manner about the Turbo mode being active? I
    have the distinct impression their posts are designed purely to
    irritate. Please don't encourage them. Anyone with any sense will
    read the article and understand the salient points you've highlighted
    about Turbo mode being an integral function of the chip.

  • Milleman - Sunday, September 13, 2009 - link

    I would say that i5 750 and Pehnom II X4 965 is fully comparable. AMD just have to adjust the pricetag and the price/performance will be on par. Looking at the Gaming rig performance, both i5 750 and Pehnom II X4 965 are well enough for gaming pleasure. I wouldn't shell out my bucks for the more expensive Intel top models. It's such a waste of money, unless you are working with huge video and image editing processes.

Log in

Don't have an account? Sign up now