Sign in with
Sign up | Sign in

Battlefield 3, Frame By Frame

FX Vs. Core i7: Exploring CPU Bottlenecks And AMD CrossFire
By

Though we usually talk about average frames per second, an even more important measure of playability is milliseconds per frame. That's because frames that take a relatively long time to render can be quite jarring. In theory, a 91 FPS rate could include a single 100-ms frame and ninety 10-ms frames, and that one 100-ms frame would be what kills your experience. 

This can happen on a single-GPU card. However, the complexities of synchronizing multiple GPUs make them more common in CrossFire and SLI configurations. We covered this micro-stutter effect in Micro-Stuttering And GPU Scaling In CrossFire And SLI, and have plans to cover this phenomenon in more depth in the next couple of months.

Since an evenly-spread 20 FPS rate would consist of 20 50-ms frames, we’re using 50 ms as the cut-off for actual playability in today’s analysis. Many gamers get annoyed with frame intervals far shorter (say, 30 ms), but that isn't as likely to get you killed as it is to simply bug you.

The performance of our Radeon HD 7970s in CrossFire appears fairly similar on our AMD- and Intel-based platforms when we run at 1920x1080. Our system based on the FX-8350 encounters a couple of higher spikes, but the worst of these we see only reaches up to 40 ms.

It's worth noting that we're using Fraps to take these measurements (currently the only solution, short of capturing the output with a PCI Express-based frame grabber). Consequently, we're not representing the entire rendering pipeline. After comparing our recorded results to actual gameplay, however, we're confident that the most egregious performance interruptions are being illustrated. Moreover, we're not comparing SLI to CrossFire, so the frame-time spikes are truly attributable to each platform.

Frame times simultaneously appear more variable (the bulk of the graph is wider) and with lower variability (the largest spikes are smaller) at 4800x900. Both platforms seldom cross the 30 ms barrier, and the AMD-based machine only spikes to 40 ms once.

You'll probably want to stop at 4800x900 or dial detail settings back to the Medium preset if 30-ms and greater frame times bother you. Ultra-quality details at this super-high resolution appear barely playable.

Ask a Category Expert

Create a new thread in the Reviews comments forum about this subject

Example: Notebook, Android, SSD hard drive

Display all 166 comments.
This thread is closed for comments
Top Comments
  • 49 Hide
    A Bad Day , January 24, 2013 3:24 AM
    Quote:
    We were hoping that AMD's Piledriver update would break that trend, but even a handful of impressive advancements aren't enough to match the effectiveness of AMD's graphics team. Might Steamroller be the evolutionary step forward needed to unleash the GCN architecture's peak performance?


    I disagree. What's needed is even stronger push on the developers to use more than four cores, effectively, not some 100% load on one core and 10% on the other five cores.
  • 34 Hide
    Someone Somewhere , January 24, 2013 4:03 AM
    Personally I'd like to see the i5-3570K included in here. It's closer in price to the 8350, but should perform more like the 3770K (as the games are unlikely to use more than 4 threads).
  • 26 Hide
    esrever , January 24, 2013 4:09 AM
    There should be an i5 included just so you can have a middle ground.
Other Comments
  • 49 Hide
    A Bad Day , January 24, 2013 3:24 AM
    Quote:
    We were hoping that AMD's Piledriver update would break that trend, but even a handful of impressive advancements aren't enough to match the effectiveness of AMD's graphics team. Might Steamroller be the evolutionary step forward needed to unleash the GCN architecture's peak performance?


    I disagree. What's needed is even stronger push on the developers to use more than four cores, effectively, not some 100% load on one core and 10% on the other five cores.
  • 1 Hide
    acktionhank , January 24, 2013 3:50 AM
    Great article and very informative. The FX-8350 really held it's own until it came down to Skyrim.

    A Bad DayI disagree. What's needed is even stronger push on the developers to use more than four cores, effectively, not some 100% load on one core and 10% on the other five cores.


    I thought more cores were for multi-tasking, as in having multiple programs running simultaneously. It would suck to turn on BF3 and everything else running on my PC simply shut down because the CPU is under 100% utilization. How would i be able to play BF3 while streaming/playing some HD content on my TV that's hooked up to my same computer.


  • 17 Hide
    alidan , January 24, 2013 3:58 AM
    acktionhankGreat article and very informative. The FX-8350 really held it's own until it came down to Skyrim.I thought more cores were for multi-tasking, as in having multiple programs running simultaneously. It would suck to turn on BF3 and everything else running on my PC simply shut down because the CPU is under 100% utilization. How would i be able to play BF3 while streaming/playing some HD content on my TV that's hooked up to my same computer.


    single core performance... look up some other benchmarks, where they use itunes to encode things, or when i believe winzip went from single core to multicore, it shows a GREAT difference more cores can do to performance.

    the problem is that few games and few programs really scale, sure, pro applications almost always take advantage of whatever you put in them, but consumer, different story.

    more cores can offer more multitasking, but they also allow the load to be shifted from one core to all 4 cores and get over all more performance when properly coded.
  • 34 Hide
    Someone Somewhere , January 24, 2013 4:03 AM
    Personally I'd like to see the i5-3570K included in here. It's closer in price to the 8350, but should perform more like the 3770K (as the games are unlikely to use more than 4 threads).
  • 4 Hide
    Crashman , January 24, 2013 4:06 AM
    A Bad DayI disagree. What's needed is even stronger push on the developers to use more than four cores, effectively, not some 100% load on one core and 10% on the other five cores.
    I'm calling BS on this one because AMD's "eight cores" are actually four modules, on four front ends, with four FP units. Games have historically been limited by FP units specifically and front ends in general, no? What I'm seeing is that Intel's per-core IPC appears to be a little higher, when two different FOUR "full" CORE processors are compared.
  • 26 Hide
    esrever , January 24, 2013 4:09 AM
    There should be an i5 included just so you can have a middle ground.
  • 1 Hide
    amuffin , January 24, 2013 4:09 AM
    I'm really liking the new logo!
  • 21 Hide
    de5_Roy , January 24, 2013 4:44 AM
    like the article.
    woulda liked to see how a 3570k does against the fx8350 running the same cfx setup. impo, the price/perf woulda tipped further in favor of intel in configs like this.
    lastly, woulda liked some newer games like sleeping dogs, far cry3, max payne 3 in the benches instead of the ol' bf3 single player. i hear bf3 sp doesn't stress cpus that much. may be bf3 skewed the benches in favor of amd as much as skyrim favored intel. :whistle: 
  • 6 Hide
    abbadon_34 , January 24, 2013 5:47 AM
    It would be nice to see prices for components similar the SMB. Not because I can't look them up, but because the article is very price/performance oriented
  • 17 Hide
    Crashman , January 24, 2013 6:52 AM
    The_TrutherizerThe final (biased/unbiased) comparisons are the oddest thing I've ever seen. You practically have to have a degree in statistics to understand them, since for us mere mortals the maximum value is 100%. How the *** you got 115% of max performance I really do not know. What is your 100%??? In any case. On the value thing I also have to add that saying that the value of the whole system is the most balanced approach is nonsense. You compare individual parts with individual parts. If you care about value then you will hardly get the most expensive case,kbd,mouse,etc... You'll buy the best CPU and gfx card and memory you can afford and stick it in any damned old chassis that you can keep cool. Your "most balanced" approach is a crock. It's like saying that two people have close as the same test scores because the average of the test scores of their respective classes are just about the same. LOL So much of that comparison was meaningless to me. Sorry...
    OK, for mere mortals: When you make this type of calculation, the average is 100%. You have to subtract 100% in order to get a 0% average. In this case, the average was "average of all systems". Think of it like IQ (where the average also happens to be 100).

    If you could buy $4 RAM instead of $40 RAM, but the $4 RAM made your system 50% slower, would you buy it? No, because it would make your $1000 PC perform like a $500 PC.

    You can only do per-component value when you're only comparing one component. In this case, the graphics cards and CPUs were being tested as a pairing (just like the title says).
  • 8 Hide
    belardo , January 24, 2013 8:15 AM
    The i5-3570 is about $100 cheaper and would play those benchmarked games pretty much just as fast as the i7-3770...
  • 13 Hide
    m32 , January 24, 2013 8:26 AM
    Thanks for the article! I just wished all games took advantage of AMD CPU(s) threads. Hopefully AMD will get better with SteamRoller and improve it's power consumption and IPC.

    Again, I enjoy reading the article. Get ready for b!tching by fanboies.... Tom. ;) 
  • 4 Hide
    Anonymous , January 24, 2013 8:31 AM
    If the AMD FX 8350 isn't fully loaded for games ... why do you use them to compare? Because it's the top CPU in AMD's portfolio?
    Why don't use AMD FX 6x00? They are cheaper, almost 60€ in my country. You have compared AMD FX 4x00 already, but i don't see any review or article using a FX6x00 and i think it's the sweet spot for an all-in-one PC (game and work, with 8GB at least of RAM).

    Sorry for my english.
  • 17 Hide
    Anonymous , January 24, 2013 8:32 AM
    You must remember Guys that the Tech in AMD's now processors don't support the Old Optimisations 3DNOW! ETC so with certain coding techniques that still use the older stuff AMD suffers. Note that with Newer Applications, AMD Especially with OPENCL can simply whoop Intel out of the Park.

    Your processor is only as good as the Programming that supports it, and Intel pays developers to use code that supports it and that is missing on AMD's architecture.
  • 17 Hide
    Outlander_04 , January 24, 2013 9:27 AM
    Compare a $200 AMD processor with a $330 Intel processor ?

    Can you run the whole test again with a $200 intel quadcore

    and ditch the old DX 9 game engines , too?
Display more comments