Note: I know this thread is a few days old, but figured I'd post my take on it anyway.
I haven't been impressed in the slightest with the dx11 support in the WoW builds that we've seen thus far.
In fact, all I've seen is Slowdowns as compared to its dx9 counterpart, with more than adequate hardware I might add.
A simple test I conducted myself should further illustrate my point.
Each preset/configuration on the respective DX API was ran 5x with the results averaged.
These tests were ran on a freshly installed Windows 7 x64 system using the following:
i7 920 at 4.0Ghz w/ HT enabled & 6GBs of DDR3-1600 in Tri-Channel.
Nvidia GTX 480 w/ latest 266.35 drivers at stock clocks.
WoW was installed to a Vertex 2e SSD.
Originally Posted by Troll Starting Area / Preset: UltraOriginally Posted by Troll Starting Area / Preset: Ultra w/ Fair Shadows, Fair Liquid & Disabled Sunshafts
So, I have no idea where people are pulling these so-called 20-30% performance increases from.
I'm certainly not seeing it; if anything, I'm finding the inverse of that.
Last edited by Trigunflame; 2011-01-13 at 06:27 AM.
Those numbers are taken from highend DX11 cards. How faster the card is how higher the gain is.
Tom's Hardware Cataclysm DX9-11 test
Im using 2x PowerColor Radeon HD6850 (since yesterday) i was running around everything in ultra with Ok fps, do you think i would gain more fps by having SET gxApi "d3d11"?
It sais from where i orderd it that it should work with DirectX 11.
I'd like to point out my post #19 in this thread. DX11 mode for me increased framerate in some areas, decreased in some with old DX10.1 card. The more Cataclysm modifications was added, the faster the area would run. For example the untouched Winterspring was slower in DX11 mode, while completely new Deepholm was faster in DX11. I didn't re-run these tests with my new GTX460 but would assume there's similar thing happening with it.
Also your second set of data has huge logic error. Water and sunshafts are especially stuff that were added in Cataclysm client, so naturally you will not see difference or suckier difference when you turn those effects specifically off.
Never going to log into this garbage forum again as long as calling obvious troll obvious troll is the easiest way to get banned.
Trolling should be.
Xfire expects a specific version of the D3D API. If you change it, Xfire can no longer work within the game. Changing the Xfire wow configuration however to expect D3D11 however should work.
Is a card designed for DX11 and the features it provides, and done so with greater focus. (Much better at tessellation for example)
Last edited by mmoca371db5304; 2011-01-13 at 10:09 AM.
A GTX 480, is about as highend as you get aside from a GTX 580 which is only a marginal difference (around ~6-10% when clocked at the same speed).
I've seen that article and I've never been able to reproduce their results (regardless of whether I OC'd this card to 850/2000 or not).
Not everyone is going to run the client at Maxed out settings. It is a well known fact the 3 altered settings make a substantial difference in FPS. As such, those results were included in the data set as well to give a realistic idea of what someone could expect.
Ideally, the margin of FPS between the use of DX9 or DX11 API, should remain constant with all else being equal.
The fact that the FPS dropped so dramatically when each option was changed (yes, though undocumented I did change each setting one by one and monitored the FPS hit) made clear that in this area, the SunShaft, Water (The entire area is surrounded in it) & Shadows effects are used considerably. Demonstrably, this showed a vast difference between DX9 or DX11 (or in the first data set, a relatively non-existent one), the fact I wasn't in a "Catacylsm" area is inconsequential to these results.
More to my point, is that you shouldn't see so called "20% gains", and conversely "20% losses" depending on the area, in the first place.
If anything, DX11 should Match the peformance of DX9 in All areas, and make the aforementioned gains in area's that make more use of these "new" effects.
The fact that it fails to do this, is all the more reason that it Shouldn't be used as it is.
I'm assuming this is yet another reason (in addition to various bugs) that Blizzard has opted to leave it disabled by default.
Last edited by Trigunflame; 2011-01-13 at 04:45 PM.
80% FPS improvement on a 6870 on ultra at highest res and full AA, only problem is the config file seems to be removing set gxApi "d3d11" every now and then, does WoW support crossfire?
Join FDWH Gold Challenge Mode on OpenRaid and get your awesome transmogrification set. These events are for everyone, ranging from hardcore veterans to skilled casual players. No experience is required, only knowledge of tactics. Make sure to put MMO-Champion in your note!
Make sure its not just being moved.
Every time WoW quits - the config file is completely recreated. It means for example that the gxApi setting is likely to be located earlier in the config file than where you placed it. The position of it however does not actually matter, and if its being moved around it acknowledges that the game has indeed read and understood the command - since its now returning it to the text file when exiting and making sure that the command remains when remaking the config file.
Last edited by mmoca371db5304; 2011-02-07 at 08:21 PM.
I got really annoyed with the WoW launcher constantly overwriting my desktop WoW shortcut (that has the -d3d11 tag to force it on) - simple answer? Set the tag, and then set the shortcut as read-only.
I to have Dx11 card, and get nearly 0 FPS with DX 11 on, yet awsome FPS with DX10 on.
lets me use d3d10 lol.. id hardly call 0 FPS *supported* either. How do i know its even worked ....
"C:\Users\Gary\Desktop\World of Warcraft\Wow.exe" -d3d11
will make it run DX11 right?
Last edited by mmoc796e664f5d; 2011-02-08 at 12:56 AM.
the client probably overwrote it with d3d9ex then, WoW never has been documented with DX10 support.