New card, new tech, new drivers, various testing methods. There is no way that this was going to be a smooth roll out.
i agree, too big mess with drivers and setups;
if they wanted to show how it would have behave in dx12 env ~ they should include mantle (dx12 should be very much alike with performance to mantle) benchmarks from dragon age, bf4 and civ.
there could be big bottleneck as furyx/hbm is built for speedy transactions which is going to slow down or flood gpu with unnecessary layers of useless data to process.
derp guys, some1 actually did it...
DX11
Mantle
Mantle v DX11
source: http://semiaccurate.com/2015/06/24/overclocking-amds-fury-x/
we're back in game.
1 Like
Still more interested in seeing some numbers regarding vram use and the effect of system memory speed on performance.
@Logan I think a video is in order. Maybe it is time for Pistol's rig to get an upgrade? ;)
I have no doubt there will be a video from Tek Syndicate about this card.
Eh hem... nudges video your way
EDIT: Whoops I was tired and this was a bad link, leaving it intact for the lulz. Here is the correct one.
Only barely beats the 980? So it matching the TitanX must be a mistake then.
1 Like
yes thats an significant diffrence, which can make the Fury X an interessting card for you guys.
In my country its not an interessting card atm.
But ofc its all still waiting for better drivers, and DX12 etc.
But the Fury and the Nano will definitely be interessting.
It's still too early to say for my country, currently we only have MSI(just arrive) and Gigabyte G1 GTX 980TI both being at the same price. These are the more expensive options for GTX 980TI, Leadtek, Palit, Asus and Zotac have yet to appear so we won't know what's gonna happen.
Sold my amd stock today; earned 30cents / share.
I was not the only one I assume someone big sold their stock - stock they brought in last week making humongous profit of 61.08Mx2.5$ = $152.7Millions I'd like to thank to AMD You are company I can count on to make profit. :*
(selling 79.44M shares brought down the price today by 30cents ~ though only temporarily, i assume it'll rise again on Monday by few cents - and once shops or themselves publish their sales.)
Though they didn't reach projected (by me) stock for ~3$ there's still hope after win10 release, and mantle benchmarks.
~
Funny part, when witcher 3 was released in may a lot of people sold nv stock (28M shares) with batman their shares went down again, might be hard time for nvidia. Though they have large cushion of money. Though their stock is not as attractive - as its hard to predict the value. I thought their shares would rise up as dx11 furyx didn't beat 980ti in most benchmarks but that wasn't true it seems.
Well I ordered the Saphire one will be posting KVM benchmarks when it comes.
cool, don't forget to test mantle with BF4 and dx11 at win7
:3
i want to see more benchmarks for dx12 alike api's.
But I only have a windows 10 VM
Mantle seem to have problems on Win10 (from what I've read), win10 ain't ready yet + runs slower by few fps in all games vs win7. Thats why i ask for win7.
Guess I could procure an ISO like I said this will be over KVM so results may vary from bare metal. Sometimes KVM gets higher FPS
what are you using for your kvm base? (outside of this conversation)
Opensuse
cool, hows the performance ~ and resource control?
Or HardOCP, Or Techpower, Or OC3D... or really any benchmark really.
Here have some links
@Phyrce
Problems is that everyone picked different drivers, and some picked wrong getting crappy performance. Currently there's only 1 mantle (closes we can get to dx12 performance) benchmark.
Bench from mantle
AMD’s Radeon Goes Premium with the Fury X
shows mantle performance much greater on fury x. In civ5 beating titan x to ground, and in bf4 would get pretty much same score as gtx980ti.
Straight from the HardOCP Bench:
"D3D 11 (DirectX 11) or Mantle
Since Battlefield 4 also supports the Mantle API we ran tests in both Mantle and DX11 to find out which API provided the highest performance above. In our testing it was determined that DX11 provided the best performance. These are the results we got in Mantle, just for comparison sake.
2560x1440 4X MSAA Ultra (First Graph Above)
Mantle Average FPS = 55
As you can see in the top graph above in DX11 mode we got 66.5 FPS, therefore DX11 was a lot faster.
2560x1440 2X MSAA Ultra (Second Graph Above)
Mantle Average FPS = 65.2
As you can see in the graph above with DX11 it was 71.4 FPS with DX11.
Therefore, on the AMD Radeon R9 Fury X D3D 11 mode is faster in Battlefield 4. This is the API you should be playing in if you want the best performance."
don't get me wrong I want the Fury X to be better, gamers deserve more from AMD for their money. I am really hoping driver tweaks and DX12 provide the bump in performance akin to the 290X. When the 290X came out it was barely competing with the 780TI, Now its usually a little higher. I want to see that performance increase sooner rather than later though.