Crysis 3

Still one of our most punishing benchmarks, Crysis 3 needs no introduction. With Crysis 3, Crytek has gone back to trying to kill computers and still holds “most punishing shooter” title in our benchmark suite. Only in a handful of setups can we even run Crysis 3 at its highest (Very High) settings, and that’s still without AA. Crysis 1 was an excellent template for the kind of performance required to drive games for the next few years, and Crysis 3 looks to be much the same for 2014.

Crysis 3 - 2560x1440 - High Quality + FXAA

Crysis 3 - 1920x1080 - High Quality + FXAA

Crysis 3 - 1920x1080 - Medium Quality + FXAA

Crysis 3 ends up being another lateral for AMD, with the R9 285 and R9 280 virtually tied at 2560x1440 and 1920x1080 with High quality settings. Only at 1920x1080 do we see them pull apart, with the R9 280 taking a mild 4% lead. On a side note, since this is our game of choice for measuring power consumption, this is especially handy as it means we have equalized performance from the start, making power consumption at the wall less affected by performance differences.

Meanwhile as this is a game that generally favors NVIDIA cards these days, it comes as little surprise to see the R9 285 trailing the GTX 760 by 4 to 8%. The factory overclocked version of this card fares a little better in that respect, but it’s not quite enough to close that large of a gap.

Battlefield 4 Crysis: Warhead
Comments Locked

86 Comments

View All Comments

  • Samus - Wednesday, September 10, 2014 - link

    Am I missing something or is this card slower than the 280...what the hell is going on?
  • tuxRoller - Wednesday, September 10, 2014 - link

    Yeah, you're missing something:)
  • Samus - Thursday, September 11, 2014 - link

    BF4 its about 4% slower.
  • Kenshiro70 - Wednesday, September 10, 2014 - link

    "something of a lateral move for AMD, which is something we very rarely see in this industry"

    Really? Seems like the industry has been rebadging for the last two release cycles. How about starting to test and show results on 4k screens? 60Hz ones are only $500 now, and that will put a little pressure on the industry to stop coasting. I have no intention of spending money on a minor bump up in specs. Bitcoin mining demand can't last forever.
  • tuxRoller - Wednesday, September 10, 2014 - link

    Error page 4: chart should read "lower is better"
  • jeffrey - Wednesday, September 10, 2014 - link

    Saw this too, the Video Encode chart Ryan.

    Congrats btw!
  • Ryan Smith - Thursday, September 11, 2014 - link

    Whoops. Thanks.
  • yannigr2 - Wednesday, September 10, 2014 - link

    Great article and review.

    The only bad news here I think is Mantle.

    But on this matter, about Mantle, maybe a slower processor could show less of that performance drop or even keep performing better than DirectX11. Maybe one more test on a slower core, on an FX machine?
  • mapesdhs - Wednesday, September 10, 2014 - link

    What were AMD thinking? How can the 285 be a replacement for the 280, given its
    reduced VRAM, while at the same time AMD is pushing Mantle? Makes no sense at all.

    Despite driver issues, I'd kinda gotten used to seeing AMD be better than NVIDIA in
    recent times for VRAM capacity. A new card with only 2GB is a step backwards. All
    NVIDIA has to do is offer a midrange Maxwell with a minimum 4GB and they're home.
    No idea if they will. Time will tell.

    Ian.
  • Alexvrb - Wednesday, September 10, 2014 - link

    There you have it, and with no issues with boost. Sorry Chizow, so much for that. :P

    Thanks for the in-depth review, Ryan. It appears that power consumption is going to vary from implementation to implementation. Lacking a reference model makes it tricky. Another review I read compared Gigabyte's Windforce OC 285 to a similarly mild OC'd 280, finding a substantial difference in the 285's favor.

Log in

Don't have an account? Sign up now