obviously is not. I never thought that a recent graphics card, so strong performance problems with such a graphically simple game has.
what are your clock speeds paarma? Reference stock or some manufacturer overclocks applied on top by default?
Problem is that AMD is pretty much better in everything else... I don't just use rF2 but I use loads of different sims & games and I have to say it's probably rf2 that will suffer for it in my case. They're cheaper and perform just as good or better in everything but here so why would I spend 500€ (asus 780D2oc) on a 780 when a 400€ (asus 290 reference) card does pretty much exactly the same and is of a newer generation with 4GB 512-bit memory, a renewed Crossfire system and Mantle support... And at the moment still has to get driver optimizations etc I agree they have a bit of issues with their coolers at the moment so but non-ref cards will surely take care of that. But regardless, I'm keeping my fingers crossed and I hope ISI crack that code
[MENTION=1]Tim Wheatley[/MENTION] Can you split all the benchmarking stuff off into its own thread? Valuable info and would be a shame for it to be buried and lost
This is my fraps bench mark test RF2 all max AA8 AF16 NO CCC control All video control via the game. Clio at Loch Drummond Frames Time (ms) Min Max Avg 221631 3172749 0 117 69.855
Don't think you have done the test right Whitmore. Your min is zero. Nobody else is. In any case you would be the only person satisfied with the performance of their AMD versus an equivalent NVidia in this software title. 60+ fps in ALL tracks is a pretty bold statement too.
The idea is to do the test in post #52 using the replay and settings stated rather than whatever you normally use. The test takes around 200 secs, your result for whatever you were testing lasted nearly 53 minutes!
Ran test as per post. Frames Time (ms ) Min Max Avg 8635 202271 21 90 42.69 When racing I do not use reflections or shadow blur.
Did you have these off for the test to ? I have a 7970 and followed settings exactly as in instructions from download and I got 27.8 average fps and previous guy had 28 in list so seems pretty consistent be interesting to find out if anything you have is different. Im like you also though always have them off and most high even with triples res I can almost achieve 60 fps consistently.
Here's my results with i5 2500k@ 3.8Ghz, Leadtek GTX680 4GB stock clocks. Frames: 13185 - Time: 200196ms - Avg: 65.860 - Min: 53 - Max: 89
Oops sorry thought I had turned them back on thought wrong. So just below you seancleric Frames Time (ms) Min Max Avg 5589 211647 14 48 26.407
Frames: 6265 - Time: 199884ms - Avg: 31.343 - Min: 22 - Max: 43 CPU: AMD Phenom II 955BE (stock) GPU: Sapphire 7850HD OC (oc to 1050/1450) Drivers: AMD 13.8 beta (and using Radeon Pro) Settings: 2560x1440 8AA/8AF (Quality) SMAA rocks!
Shamrock could you test with 1920x1080 to compared with all results in same resolution please? Thanks I updated the graph with 680 4GB Results
Yes, I agree on that. I know it's not a comfortable situation for Ati/AMD GFX cards owners but it is how it is and it's like that since rFactor 1 so while trying to not be too pesymistic I don't think, anything will change in that regard in the near future. You can always reduce graphics settings in rF 2 to have decent fps while enjoy playing in other games without such problems.
Yes i have also no hope it will change that much. It might be possible if the Mantle API could be used but i doubt. Another question, was it usefull to take a bench with max low settings on a 1920x1080 single screen ? I'm intrested in the raw performance from diff hardware with rf2. I would like to know what happens with one car and with a grid of 20 AI on low settings. What did you guys think, does it make any sense ?
Hmmm...i tought it could be usefull to see the engines basic performance. Here the first run with the Clio replay bench - everything on low -1920x1080 in use Frames-24074 , Time (ms)-200293, Min-69, Max-182, Avg-120.194 Single car Clio: Frames-42576 , Time (ms)-193214, Min-197, Max-260, Avg-220-357
That only make sense for guys with the same GPU or at least for similar GPU results like for example GTX 670 OC/GTX 680/GTX 770. Other thing that can be done in that case is test 1024x768 @ low settings with and without OC in CPU. Otherwise we won't be able to compare anything. In my case for example i have tests with my Q6600 @ 2.4/3.0 and 3.6ghz and i only notice increase performance in 1680x1050 with benchmark settings. If i move to triple screen i have same average with all 3 frequencies.
Hmmm...ok i don't know much about these things, but it Looks like as if the most performance lost is in the gfx settings and effects, not the physic engine itself. Well but even the low settings performance seems a bit to low for me is why i was intrested in testing a poor base without heavy loads on the gpu. Another intresting fact is that the 290 is really silent in rf2 in comparison as example to AC where it starts to simulate a wind tunnel.