gamegpu.ru sucks
Here are real test with max settings
No, they don't. Did you even read their review? Real test? What does that mean testing the game at cinematic console fps blindly maxing out every useless GameWorks setting for no IQ gain? GameGPU has that covered too if you actually read the entire review.
They tested UQ at 1080P only to find that FPS are in the gutter and the IQ increase going to VHQ to UHQ is almost non-existent.
37-44 fps on a 980Ti @
1080P UQ << You suggest someone with a $650 GPU should play like that?! >>
Then they actually changed the settings to simulate what a real PC gamer would use - get 99% identical IQ with double the performance and they got this:
980Ti is now in the
46-50 fps range @
1440P without using useless GW's Ultra gimmicks. Let's see, what's better playing this game at UQ with a micro-scope at 1080P @ 37-44 fps OR playing it at 46-50 fps with VHQ @ 1440P? GameGPU actually got this right on.
GameGPU actually tested the game at UQ which shows results even lower than yours. This is what you called "Real test" earlier? Ya, cinematic console fps on a 980Ti SLI. FTW!
That means if you want to play the game at UQ everything @ cinematic console FPS, GameGPU has the data for you. If you want to play the game like everyone else would - i.e., with 99% IQ and double the performance, GameGPU has that too.
Why would a game have almost no IQ increase moving from VHQ to UQ, especially for textures? Simple, because every single AA method used in this game results in Vaseline/soap smearing over your eyes. What's the difference when everything looks blurry as F. GameGPU happened to cover this as well.
FXAA = garbage
MSAA = garbage
TXAA = garbage
AA off = aliased garbage
So please explain how GameGPU sucks vs. PCLabs that blindly threw every setting to the max and reported to us that the game looks like crap and performs like crap but I guess PCLabs with their constant desire to show NV cards winning (like in 95% of games they've ever tested) didn't think outside the box, right? Contrary to PCLabs that chose FXAA+UQ for no real benefits, GameGPU actually provided
different testing scenarios to simulate how we might play the game.
Ironically, under the settings most gamers will use, AMD cards perform well relatively speaking since this game is still an unoptimized pile. I guess PCLabs is more interested in showing console level cinematic experience vs. real world gaming scenario or they didn't want to show AMD cards beating NV?
The GTX980 purchase rewards again. And no wonder the GTX970 is the best selling card far ahead of everything else on steam.
Maybe AMD performance will improve whenever they release a game ready driver.
Hilarious considering R9 295X2 is 56% faster at 1440P achieving 64 fps averages vs. 41 fps avg for a 980 under the same settings.
Yup, 980 sounds like it was a great investment.
Astoundingly low performance. All the more reason for the assassins creed series to finally die out...
At the same time the popular franchisees are their cash cows. If the major studios like EA and Ubisoft cannot get millions of more or less stable and guaranteed sales, they are going to be even less likely to invest into new directions that have more risk. Think about it, Watch Dogs can bomb because AC sells well. I agree with you though that from a consumer stand-point, this is pretty horrid. For them as a business though they need cash cow franchisees to be able to take risks. This is bad news for them since they are going to need to come up with something soon as AC Syndicate is shaping up to be one of their lowest releases in terms of sales for this franchise.