Wow wasn't expecting that sort of result from a HD7970, that's worse than my GTX470 and no increase on a CPU o/c either, it's like it's totally GPU limited. I don't get any increase going from 3.0 - 4.4GHz either but then I kind of expected my card to be totally GPU limited at these benchmark settings.
Is it possible NV has an advantage in rf2 because of Mulithreading and physx ? Ridiculous results: Phenom955BE@4ghz 1920x1080 Gigabyte R9 290 Corsair 8gb ddr3 1600mhz Frames, Time (ms), Min, Max, Avg 6923, 199992, 18, 65, 34.616 This AC, just with one car but everything max except fxaa: Frames, Time (ms), Min, Max, Avg 12009, 144244, 64, 105, 83.255 Need to test a single car around silverstone on max settings in rf2 still.
Andregee suggested AMD users use the 12.11 drivers +12.11cap2 for better performance here : http://isiforums.net/f/showthread.php/16127-triple-screen-UPDATE?p=220460&viewfull=1#post220460 Worth a try as your fps can't get much worse.
rF 2 use 2 CPU cores and here GPU doesn't make any difference. PhysX is not used either. It's all about the way nV and AMD cards cope with the way ISI engine creates a 3d scene.
Ok thx, good to know. How bad is that, just to use two cores and 32bit for such a complex thing In relation to AMD and the 290's performance i just can say, every other application i tryed so far performs great, even with the current config as a base for the 290. The noise level in idle with 20% fan speed the card is quite silent but under load the fan is pretty good to hear It's really a good gpu. I flashed the second rom with the 290x bios already and it runs fine with stock 290x clocks. The vram performs with 5500mhz atm, the gpu i left at stock 1000mhz. If the rest of my current pc configuration was on the same level,...... unbeatable R9 290 Standard with my config View attachment 10752 Here a part of a review View attachment 10753
Try the 12.11 Beta DRiver with the 12.11Cap2 Profile and you will get much more better results. all newer AMD Drivers are poor for Rfactor2 MY results. 2600k stock 7970 1175,( nearly the same like 1050mhz) min 41/ max 82/ avg (i have to guess because fraps Shows me somethin crazy maybe i am to stupid for it. ) somewhere around 50-55 . Its much more better than yours only because the 12.11 beta with 12.11 cap Maybe someone can tell me how to handle fraps for correct results All AMD users have to contact AMD like me. Till now nothing happend because one voice isn´t enough. That is a really poor Support. Ok the battlefield Gamers could be happy but whe are loosing power with new Drivers and no one cares
And this is why i stopped using ATI after my 5870...you should need no voice. It's along the same reasons for why i no longer use SLI because although when it works properly i produces bloody marvellous performance for what you pay, it was causing stuttering issues in rf2 without "vsync video" (and also at certain points in certain tracks) and i wasn't willing to go back to "vsync video" when KeiKei made me aware of the 200-300ms input lag it added. I used to be all about the highest possible numbers for the best price and stuff regardless of compromises here and there but all my prior experience of this has pushed me through and out of that phase and now I just want the least amount of hassle across the board (for what games i personally play) whilst still getting very decent performance and still at a great price/performance ratio. That's what i find with the mid-high range single nvidia gpu's (i.e. GTX 670, 770, etc).
I feel for everyone with AMD cards who's focus is on rf2 for gaming. If these performances were rectify and correlated with the performance scaling seen in all almost all other games, i would likely buy an AMD card first thing the next day.
Isn`t directx a universal API? I never worked with directx, but if you use Core Profile in OpenGL 3-4, you should expect similar performance from any vendor, not a 50% difference anyway. Only if you use nvidia/amd specific extensions, which are optimized for one of them or use exclusive features, then performance may vary significantly. Maybe rF2 engine uses some commands which are only possible to speed up on nvidia hardware (not talking about physx here)... I may be totally wrong of course and directx is not universal, and AMD/Nvidia tunes by hand driver for every app there is, but thats sounds silly. How many games gets released every day, its looks impossible to tune driver for every game just by numbers. I understand getting 10-20% in some A+ titles is possible "by hand". Its just so not typical to see 50% difference. I read hardware reviews regularly and never seen such a difference in any title (maybe one-two extreme cases 1-2 years ago).
Wait when Bf4 get the mantle API patch what the 290 series do with the NV's still on dx while even now without it, the performance is great. Only negativ i see atm is the stock cooler and rf2 performance. I see me looking at the header in 6 month still without to be able to drive with one of the most powerfull gpu's. Seems like i need to have fun with it elswhere until the guys at ISI keep there heads up
I just want to say my system altho overclocked hits something like 10800 back in July when I tested last with my GTX780, I sinced removed the app due to HD space but I will reinstall and do a screenshot when I can. My GTX780 was one of the first non reference Titan beaters. I find it funny that press only really now show GTX780 Ti as faster than Titan when back in July mine was faster stock.
Sure the 780 seems to be a great one as well. In my case the phenom base is somewhat around ~3-4 years old and doesn't supports pci-e 3.0 and even a intel 4770 is twice as fast. Under these conditions the result is still more than great for me.
My results CPU - i7 2600k GPU - HD6970 driver 12.11beta11 + 12.11CAP2 + SDK2.7 Frames, Time (ms), Min, Max, Avg 8376, 200225, 29, 58, 41.419 Using the new driver 13.11beta8, lost < 30% fps... therefore returned the previous driver 12.11beta11
So, guys with AMD cards, you can try 12.11beta11 driver, looks like it has much better performance with rF2, than 13.x. I can not use old drivers, because I need latest with proper OpenGL support.
I decided to try to make a test with the new drivers amd_catalyst_13.11_betav9.2, the result is very bad Frames, Time (ms), Min, Max, Avg 4534, 199527, 13, 42, 22.724 Will return the driver 12.11beta11+SDK2.7
May anybody of ISI should check what's the diff between the drivers. It seems that the 12.11 was the last true dx9 driver and the following ones where just optimised only for newer hardware and titles with dx10-11 support. I realised this with the CF hd6950 config as well and thereafter it became only worse. Unfortunately I have not the possibility to go back from the latest.
The ideal thing was ISI contact AMD as game developer and ask them to improve drivers or at least to gve us some feedback if is possible to improve them. Right now is ridiculous the diference in performance from Nvidia to AMD in this game.
ISI, Tim, please devote your attention to this issue. Basically everybody with this years AMD drivers have 50% of possible performance.
It's had our attention, we've tried to contact, we even made contact thanks to a third-party giving us a contact past their 'outer defenses'. Only one manufacturer supports small software titles. nVidia will literally take your software (even if you are a small company, and aren't selling their cards by the thousands), run scans on everything they possibly can, and advise you. Other manufacturers it can be hard to get an email response. I would expect AMD to gradually improve through our own trial and error, but it won't ever happen as fast or easily as nVidia because they make it possible.
Returned the driver and 12.11beta11 + CAP2 + SDK 2.8 Results Frames, Time (ms), Min, Max, Avg 8468, 200103, 30, 59, 42.318 1. uninstall the driver using ATIMan Uninstaller 7.0.2 2. Restart 3. Install AMD_Catalyst_12.11_Beta11 and AMD_Catalyst_12.11_CAP2 4. Restart 5. Install AMD-APP-SDK-v2.8 or v2.7 for your system is 64bit or 32bit 6. Restart
