[gamegpu] Far Cry 4 performance

[loccothan]

On My RIG in 1792:1344 85Hz on Ultra I have 45-70 FPS (no AA) with GodRays and Fur On -> the Game performs really Well
Optimisation is Masterpiece 😀
Phenom 4.1Ghz and XFX Default 1.169v 0% POWer

 

[Actaeon]

If TXAA and Godrays are the only things I'm missing out on by not using the Nvidia settings, then it looks like I'm not missing much.

God Rays can be enabled manually. So as far as I can tell, AMD cards are only missing the branded preset.

 

[DarkKnightDude]

TXAA in FC4 is broken anyway. I've been using SMAA/MSAA2 mostly.

 

[Silverforce11]

The question is why Ubi having access to such an optimized game engine does not use it for AC instead of using their buggy turd.

 

[daveybrat]

What i thought was amazing looking at the benchmarks is how well the low-end Maxwell based GTX 750 Ti card does.

It outperforms the GTX 660 card that normally dominates the 750 Ti.

Looks like driver optimizations really do favor the Maxwell architecture.

 

[SlowSpyder]

Wow 290/X seem to be doing great!

FX cpus not being able to catch breath is disheartening.

I think my CPU will outpace my GPU in this game, no worries here.

 

[desprado]

The question is why Ubi having access to such an optimized game engine does not use it for AC instead of using their buggy turd.

Lol it is not.Go look at steam forums or google it though it is worse than ACU.Far cry 4 stuttering all over the place.


Another benchmark of Far cry 4
http://www.pcgameshardware.de/Far-Cry-4-PC-256888/Specials/Technik-Test-Benchmark-1143026/

 

[Ryanrenesis]

LOL I can clearly tell a difference between my 120Hz monitor and my 60Hz monitor beside it.

I guess I'm a superhuman 😀

But DAMN! Dat_SLI_Scaling_Doe

 

[frozentundra123456]

Amazing, not enough hate for ACU in its own thread, now it is seeping into this one.

 

[ShintaiDK]

Lets look on this game in a month or two and see if anything is changed.

And remember, if you buy, you support the behaviour.

 

[Silverforce11]

Lol it is not.Go look at steam forums or google it though it is worse than ACU.Far cry 4 stuttering all over the place.


Another benchmark of Far cry 4
http://www.pcgameshardware.de/Far-Cry-4-PC-256888/Specials/Technik-Test-Benchmark-1143026/

That's with God Rays and NV's Soft Shadows enabled.

See what they say when its maxed but with those features off:

"But we turn off all the obvious Nvidia effects and select the highest level, AMD graphics cards can clearly those of Nvidia catch up: So the MSI GTX 970 gaming now 59.7 frames per second on average and 47 provides minimal, also by MSI derived R9 290X Gaming can make good with 57.4 and 45 (average / minimum) much lost ground. In the coming PCGH 01/2015 you can find some more benchmarks on the subject of Gameworks."

I mean its to be expected that Gameworks feature hurt AMD GPUs more, nothing surprising there TBH.

 

[SlickR12345]

The human eye can see 120+ frames per second with ease, its the human brain that doesn't compute it. Its sort of like when looking at a picture, you see the whole picture from a technical perspective, but you don't notice the details on the side, you only focus on the center.

So, you are right and wrong at the same time. The eye can easily see way beyond 60fps, but the human brain won't find it smoother over 60fps as it literally finishes the images on its own. It doesn't need additional frames.

 

[toyota]

The human eye can see 120+ frames per second with ease, its the human brain that doesn't compute it. Its sort of like when looking at a picture, you see the whole picture from a technical perspective, but you don't notice the details on the side, you only focus on the center.

So, you are right and wrong at the same time. The eye can easily see way beyond 60fps, but the human brain won't find it smoother over 60fps as it literally finishes the images on its own. It doesn't need additional frames.

if you are saying that we cant perceive more smoothness beyond 60 fps then you should really find another subject that you may actually know something about. I cant believe we would even have such an ignorant debate at the end of 2014.

 

[Enigmoid]

That's with God Rays and NV's Soft Shadows enabled.

See what they say when its maxed but with those features off:

"But we turn off all the obvious Nvidia effects and select the highest level, AMD graphics cards can clearly those of Nvidia catch up: So the MSI GTX 970 gaming now 59.7 frames per second on average and 47 provides minimal, also by MSI derived R9 290X Gaming can make good with 57.4 and 45 (average / minimum) much lost ground. In the coming PCGH 01/2015 you can find some more benchmarks on the subject of Gameworks."

I mean its to be expected that Gameworks feature hurt AMD GPUs more, nothing surprising there TBH.

So it would be about right then with the 970 leading by a couple percent. 970 is still unusually ahead but it falls within the range you would expect it to be in.

I would like to see this game in a month when driver kinks have been worked out.

 

[sontin]

That's with God Rays and NV's Soft Shadows enabled.

See what they say when its maxed but with those features off:

"But we turn off all the obvious Nvidia effects and select the highest level, AMD graphics cards can clearly those of Nvidia catch up: So the MSI GTX 970 gaming now 59.7 frames per second on average and 47 provides minimal, also by MSI derived R9 290X Gaming can make good with 57.4 and 45 (average / minimum) much lost ground. In the coming PCGH 01/2015 you can find some more benchmarks on the subject of Gameworks."

I mean its to be expected that Gameworks feature hurt AMD GPUs more, nothing surprising there TBH.

Or AMD's hardware is not as good as nVidia's. :hmm:


Infraction issued for trolling.

-Rvenger

 

[Erenhardt]

Or AMD's hardware is not as good as nVidia's. :hmm:

Of course... Every one knows that hardware inside GTX750Ti is as good or faster than HD7970.

Why stating the obvious :whiste:


Infraction issued for thread crapping.

-Rvenger

 

[Silverforce11]

Or AMD's hardware is not as good as nVidia's. :hmm:

Yes indeed. AMD's hardware is not as good at processing NV's propriety Gameworks code.

At LEAST it can run it at all, unlike GPU PhysX. I guess NV realized locking out AMD on such a feature hurts them when all the reviewers compare with PhysX OFF... now Gameworks is allowed to run on AMD hardware, but crippled, just to make the comparisons possible by reviewers. 😉

I'm sure you were just as quick to point out how inferior NV hardware was when CoH used Dx11 compute for rendering & physics calculation, or early on in AMD GE games with global illumination via DX11 or even just a few days ago in Dragon Age Inq benchmarked showing the R290X pwning the 980 and the 780ti harder... wait... You didn't.

 

[Grooveriding]

So the major bugs in the game are a single black bar at the bottom of your screen at 2560x1600. The game is always letterboxed, but it should be at the top and bottom of the screen. This appears isolated to this resolution.

http://forums.ubi.com/showthread.ph...ge-black-bar-only-at-the-bottom-of-the-screen


The other is the SLI & SMAA ghosting bug introduced by nvidia with their 344.75 drivers. If you run SLI and have SMAA on you get a sort of transparent ghost image everywhere that appears as the shadowing of foliage.

http://forums.ubi.com/showthread.php/954827-SLI-Ghosting-after-Nvidia-update

Then there is the severe mouse acceleration issue.

http://forums.ubi.com/showthread.ph...-mouse-acceleration-game-is-horrible-to-play?

Guess no ubi game is ever not somewhat to completely broken. But performance in FC4 is good and the engine is optimized regardless of some issues.

 

[monstercameron]

if you are saying that we cant perceive more smoothness beyond 60 fps then you should really find another subject that you may actually know something about. I cant believe we would even have such an ignorant debate at the end of 2014.

It is all subjective so get off your high horse.

 

[Erenhardt]

Running the game higher than 30 fps is hurting the cinematic effect - their own words.

 

[96Firebird]

So the major bugs in the game are a single black bar at the bottom of your screen at 2560x1600. The game is always letterboxed, but it should be at the top and bottom of the screen. This appears isolated to this resolution.

http://forums.ubi.com/showthread.ph...ge-black-bar-only-at-the-bottom-of-the-screen

Do you have DSR enabled?

I'll try setting my resolution to 2560x1600 with DSR and see if I get the same issue.

 

[SlowSpyder]

Or AMD's hardware is not as good as nVidia's. :hmm:

Yea, that's it. A Nvidia sponsored game that uses Gameworks runs better at launch on Nvidia GPU's, must be their vastly superior hardware. Gaming Evolved titles often seem to run pretty well out of the box on both. Gameworks, not so much. [Nope]


Please don't call him out, We don't need another AMD vs Nvidia war and a closed thread.

-Rvenger

 

[raghu78]

Yes indeed. AMD's hardware is not as good at processing NV's propriety Gameworks code.

At LEAST it can run it at all, unlike GPU PhysX. I guess NV realized locking out AMD on such a feature hurts them when all the reviewers compare with PhysX OFF... now Gameworks is allowed to run on AMD hardware, but crippled, just to make the comparisons possible by reviewers. 😉

I'm sure you were just as quick to point out how inferior NV hardware was when CoH used Dx11 compute for rendering & physics calculation, or early on in AMD GE games with global illumination via DX11 or even just a few days ago in Dragon Age Inq benchmarked showing the R290X pwning the 980 and the 780ti harder... wait... You didn't.

rofl. his logic is so fragile. When Nvidia does poorly in a Gaming evolved title its a conspiracy or deliberate attempt by AMD to sabotage Nvidia performance and when AMD does poorly in a TWITMBP title its AMD's poor drivers or inferior hardware. 🙄

 

[96Firebird]

rofl. his logic is so fragile. When Nvidia does poorly in a Gaming evolved title its a conspiracy or deliberate attempt by AMD to sabotage Nvidia performance and when AMD does poorly in a TWITMBP title its AMD's poor drivers or inferior hardware. 🙄

Am I seeing this right ? R9 285 beating GTX 770, R9 280X beating GTX 780 and R9 290 beating GTX 780 Ti across all resolutions. If these benchmarks are indicative of actual game performance then my question is "Is Nvidia holding back Kepler performance (in drivers) to sell more Maxwell cards." ? I am going to wait and see if other websites benchmarks agree. But this is quite embarassing for Nvidia Kepler cards like GTX 780 Ti , GTX 780, GTX 770 especially in a TWIMTBP title. :biggrin:

And then, according to you, if AMD cards do well in a TWIMPBT title, it is because Nvidia is "holding back performance to sell more Maxwell cards".

I know you like to think you're so much better than him, but you're not. Your logic is just as bad.

 

[monstercameron]

And then, according to you, if AMD cards do well in a TWIMPBT title, it is because Nvidia is "holding back performance to sell more Maxwell cards".

I know you like to think you're so much better than him, but you're not. Your logic is just as bad.

Care to speculate rather than this meta commentary?