Discussion in 'Videocards - AMD Radeon Drivers Section' started by oGow89, Jan 27, 2016.
GCN 1.1 Only? why?
If the improvements are there for Nano it works for later ones. It is supposed to read gcn1.1+ most likely. Someone on reddit asked around already and got confirmation.
no GCN 1.0 shame amd i am so disspointed :wanker:
Thanks for the numbers. Do you happen to have raw frame data? I believe that async is much more beneficial for frame times/smoothness, than it is for average framerates.
So cool and perfect timing as brought my RX 480 today. Upgraded from AMD 7850!!!! Cant wait to reply again in 2k!
Hilbert HAS to update his benchmarks. I want to see AMD cards MURDER Nvidia cards even though it's a Nvidia optimised game.:banana:
For me GCN 1.0.
Received a massive boost in performance.
I have in this video a big amount of info with a lot of fraps logs turned into graphics
If you don't like video, I have a full excel file
Spoiler alert: A boost in the same level as the nano of our friend here
I can confirm that CrossFire is working now in DX 12 mode. However, the performance is slightly worse for me. I used the benchmark to compare.
Wait, how are you running a patch7 with gcn1.0 in the first place?
how do you enable/disable asynchronous rendering?
Just updated, started the game under dx12 and benchmarked
They only claimed that Async compute was exclusive for GCN 1.1 or above, not that I can't run the game at DX12.
Async compute active or not, the performance boost was huge comparing to patch #6 XD
Holy ****. Did NVIDIA get similar percentage increases? If they didn't, then there is NO DX12 title that NVIDIA is faster than AMD.
Would love to see that, but i just have a kepler card, I might take a look in 1 hour or two, if you want ^^
EDIT: Yep, massive improvement too on my Kepler GT 740M comparing to the patch #6
Used to had Patch #6:
DX12: 41.81FPS ^^
Why AMD? they don't make this GAME.
Besides is a tremendous work for the devs to crank up Async for All GCN's
You'll need to program render.etc. Pipes for Every GCN out there:
and Polaris 4.0
As some Devs says it has Great Potential but to unlock it You need to Code paths for every GCN separately.
Note: R280X will also benefit from this Patch, just check it out and see :nerd: (it has some shared resources in Render Path for ACE, so it will be some gain)
Yup, i just read your post lol
Read mine :book:
What is this for a broken wording?
And of course DX12 is also faster on Nvidia in CPU limited scenarios.
Awesome! Thanks for testing!
There is no "of course" for anything, unless it gets tested.
For me , this patch gave me a massive performance boost on my laptop with R9 m265x GPU which is GCN 1.0 running in Dx12 mode. Before that lot of shuttering was there in dx12 and i dont know whether its a placebo or not Image quality is improved a bit. I am on Win 10 x64 14383 build.
Any review on FuryX performance patch 6 vs 7?
Cant find one.
Search on Reddit, there is one with Nano