-
We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.
[gamegpu] Far Cry 4 performance
Page 5 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
Silverforce11
Lifer
Well it's Ubi, MSAA takes a massive performance hit in ACU as well. It looks like they didn't bother to optimize it, period.
Whereas in Frostbite, its a deferred rendering engine as is Ubi's engine for AC, but DICE worked to optimize MSAA so it incurs a moderate ~25% performance hit, not massively as we're seeing here. Now, you guys can argue all you like about brands, but I trust that DICE >>>> Ubi when it comes to game engine optimizations.
The more accurate way is that MSAA is going the dinosaurs on Ubi games, because with Gameworks pushing shader AA, they just won't take the effort to optimize proper MSAA. There's the argument that shader based AA is fine, but it's not when it blurs the game, detracting from all that wonderfully crafted high-res textures. I know Ubi thinks it's more "cinematic" to have a blurfest running at 30 fps, but certainly that view isn't held by many PC gamers. A proper AAA PC game is expected to deliver multiple optimized AA modes.
Only to validate MFAA, if you enable MSAA and do not select MFAA in the GPU management panel the card will run on MSAA, but once MFAA is activated MSAA will be replaced by MFAA, i think it s clear enough that you need it only as a validator on the management panel but that it wont be enabled on the game itself, dont understand why you re insisting that MSAA will be enabled, all the purpose of MFAA is to not use MSAA because :
I do not have the habit to troll and do made up statements on the behalf of others.
"In an obvious maneer, Nvidia prefer to see the things under the former angle and would see no objection that a reviewer compare the performances of Gforces in MFAA 2xT with the ones of a Radeon on MSAA 4x".
It is clear enough, actualy you didnt read the article and are just dodging.
That s a marketing implementation, so the gullible think that MFAA works additionaly to MSAA while it wont, all the purpose is to offload the GPU thanks to this recycled anti aliasing interpolation.
Last edited:
Im guilty of glancing through the article, because google translate is terrible.
But your claim is still a leap:
Seems like innocent speculation on author's part, even somewhat funny,
as opposed to him actually being pushed into comparing NV MFAA with AMD MSAA?
And its actually the other way. Nvidia insists on comparing unaltered, original images as envisioned by programmer, and without any driver trickery or additional optimizations.
I might find this statement, if you are as persistent as I have been with you 🙂
But your claim is still a leap:
Seems like innocent speculation on author's part, even somewhat funny,
as opposed to him actually being pushed into comparing NV MFAA with AMD MSAA?
And its actually the other way. Nvidia insists on comparing unaltered, original images as envisioned by programmer, and without any driver trickery or additional optimizations.
I might find this statement, if you are as persistent as I have been with you 🙂
Oh man, yet another game where NV seems to have completely forgotten to optimize for Kepler architecture vs. Maxwell. 970 > 780Ti and 280X is just 5 fps slower than 780 at 1080P. Ridiculous. I would be sooo pissed if I had purchased the 780 at $650 and 780Ti at $700 at launch. Jeez, the 970 is as fast as a 690.
Man, I'd be so pissed if I bought a 290 to get two measly fps over a 2980X.
Another review from pclab and sweclockers.
http://pclab.pl/art57559-6.html
http://pclab.pl/art57559-7.html
Can't tell why R9 290X performance is so different from one review to another. Much better here especially at 1600p. Is it god rays or PCSS+ or MSAA. AMD better fix the performance for those cases.
http://www.sweclockers.com/artikel/19647-snabbtest-grafikprestanda-i-far-cry-4
AMD better fix those min fps too 🙂
http://pclab.pl/art57559-6.html
http://pclab.pl/art57559-7.html
Can't tell why R9 290X performance is so different from one review to another. Much better here especially at 1600p. Is it god rays or PCSS+ or MSAA. AMD better fix the performance for those cases.
http://www.sweclockers.com/artikel/19647-snabbtest-grafikprestanda-i-far-cry-4
AMD better fix those min fps too 🙂
Last edited:
DiogoDX
Senior member
Is not tesselation because 285 is slower than 280. But the performance in the 285 is weird becuse the min fps are lower than 270X.😕
Last edited:
njdevilsfan87
Platinum Member
To be fair, FC4 actually looks GOOD with maxed out settings.
That LOOKS like it should take insane GPU power to hit and the image correlates to it.
If you took our the character model, you could almost say that's a real life photo.
The lighting looks good, but the textures (as usual with cross platform games) just look so console. I wish Ubi would allow PC modders to do their thing like they have done with Skyrim. "Modded FC4" would be amazing.
The sad part is that it's FREE for them to allow it to happen. So why not? They get more press in that regard too. How much press did Skyrim get over mods? Tons of extra articles on websites about "Look at FC4 with xyz mod!!!!" or "Look at FC4 with ENB Lighting mod x818xbestestevermod!!!!"
Just a ton more exposure.... for a company like Ubisoft trying to milk every penny from a game you'd think this would be mandatory...
njdevilsfan87
Platinum Member
You've have me sitting here like, "look at my modded FC4 using 5-6GB of VRAM on my Titans! Runs like total crap (or looks so cinematic) but looks awesome!!!"
But also, modding Skyrim gameplay mechanics alongside of graphics has turned "modded Skyrim" into such a better product than the vanilla version. I'm also surprised Ubi wouldn't want to give the PC community modding capability of a sandbox game that screams for this kind of community activity. Elder Scrolls modding capability is really what sells that series, imo. It's not like Besthesda has had any issues selling DLC either.
Last edited:
If you're referring to Battlefield 3 and 4, I already told you that they use a much less sophisticated lighting setup (mostly prebaked with few dynamic light sources) than AC and Far Cry 4. That's why the performance hit for MSAA is comparatively much lighter.
Look at Dragon Age Inquisition if you want to a more meaningful comparison. DAI has a much larger performance hit for MSAA than BF4, because it uses a type of global illumination..
Silverforce11
Lifer
Here's the cool thing about Frostbite, if you're not happy with MSAA quality or performance, you can run it via downsampling (in DAI & BF4).
Set it for 1600p and you get a crisper image on your 1080p setup, that's FASTER than running 1080p with 4x MSAA (due to the IQ & perf penalty of deferred renderers).
These features are going above and beyond the expected for a AAA PC developer. It's something Ubi doesn't give think about with their console mentality.
3DVagabond
Lifer
They need to keep us upgrading somehow. I've felt for a long time that they just continue to artificially load GPU's to kill performance.
3DVagabond
Lifer
Support for modders is not free. It can actually be a PITA for the devs to have to offer continuing support. Now, I don't agree with not doing it though. I think the biggest part of the "magic" with PC gaming is the fans turning it into their own. It's a special breed of Dev that finds the pleasure in watching their game grow and evolve further than what they had envisioned.
One more FC4 performance review is up. Ultra with SMAA. AMD makes an excellent showing with R9 290X beating GTX 970 and GTX 780 Ti. Nvidia has work to do according to the reviewer. ubisoft also has a lot of work to do to fix stability.
http://www.techspot.com/review/917-far-cry-4-benchmarks/page4.html
http://www.techspot.com/review/917-far-cry-4-benchmarks/page3.html
http://www.techspot.com/review/917-far-cry-4-benchmarks/page2.html
"Given that we found the Radeon R9 280X to be faster than the GeForce GTX 780 at every resolution, there's no doubt some controversy will follow these results.
We figured something was wrong from the get-go so we contacted Nvidia and they suggested using the newer 344.75 drivers, which were given to us prior to being made public -- initially we were using the 344.65 WHQL. The new driver was pressumably tweaked for Far Cry 4, though the release notes don't make any performance claims. Alas, with the suggested release installed, we didn't receive a single extra frame. We should also point out we are testing with the latest version of Far Cry 4 (v1.3 as of posting).
For all Radeon cards we used the Catalyst 14.11.2 beta driver, which has also been updated for Far Cry 4. AMD claims up to a 50% performance increase over the Catalyst 14.11.1 beta in single-GPU scenarios with anti-aliasing enabled.
It's worth mentioning this latest beta driver doesn't support CrossFire. The CrossFire profile for Far Cry 4 is currently disabled while AMD works with Ubisoft to investigate an issue where CrossFire configurations are not performing as intended.
Getting back to Nvidia's poor performance... we can confirm that the 344.75 driver was used while Far Cry 4 has been patched to the latest version through Uplay. We asked Nvidia if the performance we saw was unusual or different to what they have seen and they have yet to reply.
As it stands, we believe AMD is getting the most out of its Radeon graphics cards in Far Cry 4 and don't expect to see many performance improvements in the future, with the exception of CrossFire setups. Nvidia on the other hand have some work ahead, which is hard to believe with Far Cry 4 being Nvidia-sponsored."
So AMD cards perform well with Ultra and SMAA. so the gamegpu results with Ultra and SMAA were not a anomaly. AMD still has to fix MSAA and Crossfire performance. Nvidia cards do well with MSAA but with SMAA they could improve performance.
http://www.techspot.com/review/917-far-cry-4-benchmarks/page4.html
http://www.techspot.com/review/917-far-cry-4-benchmarks/page3.html
http://www.techspot.com/review/917-far-cry-4-benchmarks/page2.html
"Given that we found the Radeon R9 280X to be faster than the GeForce GTX 780 at every resolution, there's no doubt some controversy will follow these results.
We figured something was wrong from the get-go so we contacted Nvidia and they suggested using the newer 344.75 drivers, which were given to us prior to being made public -- initially we were using the 344.65 WHQL. The new driver was pressumably tweaked for Far Cry 4, though the release notes don't make any performance claims. Alas, with the suggested release installed, we didn't receive a single extra frame. We should also point out we are testing with the latest version of Far Cry 4 (v1.3 as of posting).
For all Radeon cards we used the Catalyst 14.11.2 beta driver, which has also been updated for Far Cry 4. AMD claims up to a 50% performance increase over the Catalyst 14.11.1 beta in single-GPU scenarios with anti-aliasing enabled.
It's worth mentioning this latest beta driver doesn't support CrossFire. The CrossFire profile for Far Cry 4 is currently disabled while AMD works with Ubisoft to investigate an issue where CrossFire configurations are not performing as intended.
Getting back to Nvidia's poor performance... we can confirm that the 344.75 driver was used while Far Cry 4 has been patched to the latest version through Uplay. We asked Nvidia if the performance we saw was unusual or different to what they have seen and they have yet to reply.
As it stands, we believe AMD is getting the most out of its Radeon graphics cards in Far Cry 4 and don't expect to see many performance improvements in the future, with the exception of CrossFire setups. Nvidia on the other hand have some work ahead, which is hard to believe with Far Cry 4 being Nvidia-sponsored."
So AMD cards perform well with Ultra and SMAA. so the gamegpu results with Ultra and SMAA were not a anomaly. AMD still has to fix MSAA and Crossfire performance. Nvidia cards do well with MSAA but with SMAA they could improve performance.
BFG10K
Lifer
Huh? FXAA has absolutely nothing to do with MFAA. MFAA uses a base MSAA and blends the results between frames. It's actually very similar to TXAA in many respects.
In theory the performance hit should be a little higher than the base MSAA due to the overhead of the multi-frame blending and the post-filter calculation, and in practice that's exactly what we're seeing:
3DVagabond
Lifer
Any possibility that GCN being in the consoles is helping being better optimized for AMD so soon after release?
Silverforce11
Lifer
I don't understand the poor showing from Kepler, especially the 780 and 780ti, it makes no sense why they should be so far below Maxwell.
Gloomy
Golden Member
If we're taking wild guesses, then I'd say it's probably this combined with Nvidia getting in late in development.
Judging from the bugs and totally different lighting with Nvidia's tech it's safe to say Gameworks integration was a rushjob.
ShintaiDK
Lifer
nVidia got fixed just as fast in other games that was pro AMD.
Also look at where the R9 285 sits. Down below with the HD7950.
Last edited:
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 25K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes + WCL Discussion Threads
- Started by Tigerick
- Replies: 25K
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 24K
-
-
