Live Performance Benchmarking Comparison for rFactor 2

The Titan is supposed to be slower than the 780ti

 

I don't think it's quite as simple as that, there's something going on with some of these systems that means they're not performing as they should.

 

Which is exactly what my results show. What's your point?

 

OK, I put the Titan into System 2...

Frames: 10731 - Time: 67720ms - Avg: 158.461 - Min: 125 - Max: 186

Compared with the Titan results from System 1...

Frames: 8787 - Time: 67564ms - Avg: 130.054 - Min: 100 - Max: 156

Again we see that System 2 performs approximately 25% better than the same GPU in System 1.

 

To quote myself, just installed 344.11 and can't even get into rF2. Keep getting an error that no DirectX9.0C video adaptor is found >_>

 

Oh goody, this is turning out to be a real plot twist!

 

Hello. Nice test you start DrR1pper. I want also testing to see how my older card stand up to those new. But i have next problem. My system is in my profile, i use latest radeon drivers and have instaled radeon pro, which was downloaded from your link. Problem is i can't start rfactor2 launcher when radeon pro is runing. Other games seems work fine. If i look under task manager i can see process "launch rfactor.exe" is runing but launcher windows is not appearing. I was already set the rfactor2 profile and set the launching method, switch between 32/64 bit, add the executable name to RadeonPro Compatibility list, set the administrator rights, uninstall Catalyst control center, go through every setting and combination in radeon pro i can find and nothing seems to resolve this issue. The only think that left is reinstall rfactor2 and i will definitely not do that just to run performance test. Is simply not woth the work because without radeon pro my rf2 installation working perfectly. But i will be very glad and thankful to anybody who perhabs have same experience or hear for same problems from others. Thnx.

 

Hi kimikaze, sorry but i can't really help, I don't have an ATI card. Perhaps someone with experience of your problem can lend a hand.

Alternatively, you can just check your cpu utilisation (in task manager) from during the benchmark and if any core reached 100% then you're results are very likely cpu bottlenecked. If no core reached 100%, you're safe.

 

Yeah don't worry, because probably is some stupid bug or some other problem related to rfactor.exe and radeon pro, which is not worth our time Anyway thnx for quick response but bottlenecking is really not an issue. For that the game need to be started first and i even don't see rfactor launcher window so two steps before i can start evaluate for bottlenecking So back to topic and testing results

 

I think there's an OSD option with Radeon Pro? Try turning that off, see if it helps.

 

Try closing the "launch rfactor2" thing in your taskmanager and click the shortcut again, and keep doing that untill it opens up. I have the same problem and that's the only solution.

So since I rarely reboot my PC I just leave the launcher open at all times.

 

You're welcome http://isiforums.net/f/showthread.php?p=307847

- Don't worry, it doesn't take nearly as long as it looks

People use RadeonPro just fine with rF2. Make sure you linked RadeonPro to the correct version of rF2 (64 or 32). And you uninstalled Catalyst Control Centre??... Why would you do that??... Try setting Radeon Pro to full default settings. Or reinstall Catalyst Control Centre....Are you sure you didn't uninstall the Display Driver itself or anything else vital when you uninstalled CCC? Something is not right, I've used RadeonPro with rF2 tons of times. Also don't adjust any of the setttings of RadeonPro, not even the ones in the pic DrR1pper has on his OP, just adjust one setting only - check the "flip queue size" box and set the slider for it to "1". I've used radeonPro with like 7 different AMD GPU setups, it's an amazing program and works with rF2, no problem.

 

OK, I think we've solved the problem of low GPU performance on some systems, or at least on my System 1 mentioned earlier.

The problem? The GPU was stuck at PCIe 2.0 instead of 3.0.

The cause? My system 1 has an SandyBridge-E CPU on an X79 motherboard. In order to enable PCIe 3.0 on Sandybridge-E/X79 board you need to run a tool from NVIDIA, available here: http://nvidia.custhelp.com/app/answers/detail/a_id/3135/~/geforce-gen3-support-on-x79-platform

So, you guys with lower than expected performance, are your cards also stuck in PCIe 2.0 mode?

 

Techade, I have X79 just like you...and fully use and have PCI-E 3.0 @ full 16x on both my GPUs in SLI, no fix or whatever needed.

EDIT: You're using a Sandy Bridge-E CPU, yes, I think in your case you need the fix. With Ivy Bridge, Ivy Bridge-E, Haswell, and Haswell-E you get proper PCI-E 3.0 "out of the box".

Then there's people with the non-E Sandy Bridge CPUs (2500k, 2600k, 2700k, etc.) they are stuck with PCI-E 2.0 no matter what, I believe. Sucks for them.

 

Genius! Well done mate and thanks for figuring it out!

 

Yeah and you're not the one with the performance loss. So it all makes sense.

Timpie, you need to figure out how to get your card running in PCI-E 3.0 mode on your motherboard! Then you're performance loss should be gone.

I have bad news for some people with older CPU's and fast enough graphics cards to benefit from in rf2 with PCI-E 3.0 (though i'm not sure where that cut-off line is).

If you have 3rd or 4th Gen intel Core cpu's, you're fine. If not, you're screwed like me. I have an i5-2500k and since PCI-E is integrated into the processor, my mobo can't use the PCI-E lane in 3.0 but only 2.0 (which i can confirm it is doing so).

 

He said X79 needs an update, or whatever, to enable PCI-E 3.0 and I'm sayin NO, it's not X79 that needs the update, it's the fact that his CPU is a Sandy Bridge-E CPU, they need the update, not the Ivy Bridge-E CPUs like mine (which also use X79 boards).

 

Post updated, that better?

 

http://www.pugetsystems.com/labs/articles/Impact-of-PCI-E-Speed-on-Gaming-Performance-518/

Anyway, my PCIX is running at 3.0 x8, I guess because I have a soundcard in another PCIEX slot.

Reason RF2 wouldnt start was because the hacked dll (which came from the 335.23 drivers) to force lod bias to 0 was still in my bit64 folder, and it obviously didn't work with 344.11.
Benched them aswell when I figured it out and got exactly the same as on 335.23 but on 335.23 I could use the dll hack so I went back to that

 

PCIe bandwidth might not make much different in the titles they tested, but it sure does in rF2...

GTX Titan @ PCIe 2.0
Frames: 9085 - Time: 67705ms - Avg: 134.185 - Min: 103 - Max: 163

GTX Titan @ PCIe 3.0
Frames: 10558 - Time: 67423ms - Avg: 156.593 - Min: 124 - Max: 182