Crysis 3 Performance, Benchmarked On 16 Graphics Cards

[digiex]

Rich boys game.

 

[excella1221]

I have a hand ful of benchmarks I ran when I upgraded to from the i3 3220 to the i7 3770K and I DEFINITELY noticed a jump in the minimum frame rates:

http://www.tomshardware.com/forum/395367-33-crysis-benchmark-560ti

Dry

This is actually interesting.
Hopefully, Tom's would do a CPU benchmark of Crysis 3. Maybe the Piledriver series does good here.

 

[JOSHSKORN]

Where's the GTX 680 SLI?

 

[de5_Roy]

This is actually interesting.
Hopefully, Tom's would do a CPU benchmark of Crysis 3. Maybe the Piledriver series does good here.

there is a cpu benchmark in the article. it doesn't have core i7 3770k but pildriver cpus are there - core i3 class smoothness with core i5 class fps.

@cleeve: in the cpu benchmark chart, the llano apu is labelled as 'regor', core i3 3220 is called 'sb' - aren't those 'husky'(and quad-core) and 'IB' respectively?

 

[beetlejuicegr]

I have to whine a bit too.
Many of the users are still using sandy bridge cpus. To be honest the word "still" shouldn't even be there.
Many of the users are still using GPUs on the range of nvidia 4xx, 5xx and amd 5xxx, 6xxx.

Is there a reason for not throwing in some hardware like this on new heavy games tests? Is it for advertising new hardware? Or just lack of the harware at your test rooms? If it's the latter i will gladly send you my hardware (you can send me the newest because you always have some there ^^ )

 

[pauldh]

[citation][nom]beetlejuicegr[/nom]I have to whine a bit too. Many of the users are still using sandy bridge cpus. To be honest the word "still" shouldn't even be there.Many of the users are still using GPUs on the range of nvidia 4xx, 5xx and amd 5xxx, 6xxx. Is there a reason for not throwing in some hardware like this on new heavy games tests? Is it for advertising new hardware? Or just lack of the harware at your test rooms? If it's the latter i will gladly send you my hardware (you can send me the newest because you always have some there ^^ )[/citation]
If I may jump in here. I totally understand and share your desire to add more and more and more hardware. But you need to understand the lab time that Don already put into a story like this. 16 graphics and 9 CPU variations made it into this story, besides all the other coverage. Lines must be drawn somewhere, to get the story out timely, while keeping drivers fresh and assuring the data remains the most relevant to you. You could always make some gentle suggestions for specific hardware that you think adds value. Tom's values your input!

I like multiples... multiple CPUs tested with multiple GPUs, and at multiple settings... the more the better. But that is a dream, it quickly becomes unmanageable, and even on it full-time no one person could keep up with driver and game updates. Then you need to chart it all, without making people's head spin. =)

 

[blppt]

[citation][nom]JOSHSKORN[/nom]Where's the GTX 680 SLI?[/citation]

The 690 is close enough to a 680 SLI that you could pretty much take its results.

 

[ohim]

Seeing this benchmark just shows that going AMD FX 8350 you have a lot of spare cash for a better graphics card and get better overall performance, but all this unless you sit on huge sack of money.

 

[Guest]

Running on gtx 560 in sli on an i7 rig and it runs high without issues.
Crytek have also implemented 3d properly so it doesn't run thru NVidia's 3d system (thereby limiting it to 720 resolution). Running it at 1080 in 3D projected on an 8ft screen is awesome!

 

[mayankleoboy1]

[citation][nom]pauldh[/nom] You could always make some gentle suggestions for specific hardware that you think adds value. Tom's values your input![/citation]

I would politely suggest a Intel Core2Quad, clocked to a respectable 3.3 GHZ.

 

[mayankleoboy1]

I have yet to play a game as awesome as Crysis Warhead in the Crysis series.

 

[realibrad]

Please fix the 670 graph on page 2. Minimum cannot be higher than Avg.

Hopefully we see more games that use more than 1 core.

[EDIT BY CLEEVE: Chart fixed, thx! /EDIT]

 

[Guest]

There is no way this game should bring current rigs " to their knee's" it looks great but not for the performance demand it brings. I mean seriously if anyone thinks about it, they got this to run on the xbox 360??

Yet you want me to believe I need a 1000$ worth of current graphics cards in my system to run this smoothly for a small bit of sharper graphics? They totally artificially inflated the requirements to boost graphic card sales, period.

 

[TheMadFapper]

GTX 670 is obviously the best value pick out of all of these, very happy to see I can play this game quite easily with just one.

 

[Memnarchon]

Hello.
Someone (Ryan Shrout) at PC Perspective made an interesting test, changing a bit the way he measures the frame time percentiles and he noticed that the performance metrics that the industry is gathering using FRAPS are inaccurate in many cases and do not properly reflect the real-world gaming experience the user has.

Is there any possibility Tom's Hardware will do the same on future tests?
Thanks in advance.

 

[tlg]

A single 7970 at 1000/1400 handles MSAA 4x perfectly fine as long as you disable Vsync. You're not going over 60fps anyway so you dont need it.

 

[tlg]

[citation][nom]excella1221[/nom]This is actually interesting.Hopefully, Tom's would do a CPU benchmark of Crysis 3. Maybe the Piledriver series does good here.[/citation]

It is actually known that Crysis 3 uses Hyperthreading and if have a very possible impact on your FPS

 

[Cryio]

"We want to show you the difference between quality settings, so let me post these compressed-as-hell jpgs/gifs for your pleasure."

 

[FinneousPJ]

lol there's a problem with this graph

http://media.bestofmicro.com/A/M/374926/original/AA-Perf.png

obviously avg fps cannot be lower than min fps

[[EDIT BY CLEEVE: Fixed, thx! /EDIT]

 

[ojas]

Since FXAA and 2x SMAA are compatible across all of the cards we're testing, these are the modes we'll focus on in our benchmarks.

MSAA?

EDIT: i mean, that's compatible across all cards too...

 

[darkchazz]

[citation][nom]mouse24[/nom]Wonder if theres tessellation under the ocean in this one to.[/citation]
This argument has been debunked by a Crytek dev. C2 does not render tessellated ocean under the ground unless you switch to wireframe mode, which is what the clueless journalists did.
http://www.crydev.net/viewtopic.php?p=889047#p889047

 

[ojas]

[citation][nom]Immoral Medic[/nom]I completed this game in 4.5 hours. I gotta say, having great graphics does NOT make a good game. It's sad when all you have to attract customers is "Best Graphics in a Game Yet". BUYBUYBUY. Don't even get me started on the absolutely terrible multiplayer...[/citation]
Yeah, was reading the RPS review, than panned it.
http://www.rockpapershotgun.com/2013/02/21/crysis-3-review-pc/

Hope i don't seem like a spam bot :lol:

 

[darkchazz]

[citation][nom]stickmansam[/nom]Still feel that the game is unduly harsh for what it displays[/citation]
Then you haven't seen it in person yet.

 

[arkadi]

I have ASUS g55 laptop connected to 60" 3dtv...the game looks nice and it is playable @720p

 

[cleeve]

[citation][nom]Memnarchon[/nom]Hello. Someone (Ryan Shrout) at PC Perspective made an interesting test, changing a bit the way he measures the frame time percentiles and he noticed that the performance metrics that the industry is gathering using FRAPS are inaccurate in many cases and do not properly reflect the real-world gaming experience the user has.Is there any possibility Tom's Hardware will do the same on future tests?Thanks in advance.[/citation]

FRAPS is only inaccurate for multi-card setups.

We actually have the same hardware as PC Perspective now, we're using in in a in-depth analysis of the problem in the near future.