It is wasted time mate he just don`t get that 1080Ti or any other card > PCIe@2.0x16 ~5% loss at most under worse cases in other games but not in rFactor 2.
Yep. According to my tests PCI-E starts to bottleneck clearly at 30% usage and significantly when above 50%. It doesn't indeed ever go to 100% even on PCI-E 2.0, but it's still a bottleneck.
Synthetic benchmark at the end. For rFactor2 and R3E switching from PCIe 1.0 to 2.0 did a HUGE difference in performance for me. Interesting it did made no difference on AC, so it definetly have to do on how the rendering engine is designed. That tomshardware video is totally useless and uninformative for this context. I do not know how much difference it would be from PCIe 2.0 to 3.0. One way to possibly check could be using Rivatunner stats and see the % usage of FB and BUS.
not on all systems I tested on mine and PCIE 2.0 vs 3.0 drop with 1080ti is ~7%. Anyway, why run in 2.0 if you can run in 3.0? I also tested CPU downclock/overclock and got linear growth/drop in FPS with frequency raising/dropping. But as stonec said, I only tested with full grid visible, because that's the scenario I care about. It doesn't matter to me what FPS I get with no vehicles visible. I am building 8700k system, might try 2.0 vs 3.0 for fun, but probably now I'll see drops you're seeing.
How are you getting that conclusion? Not questioning you are wrong but I think I am close to the same result.
What are your system specs? Asking because as I got close to 50% I start to see fps downgrade with no full usage of both CPU/GPU, which tells me they are being held by a bottleneck which probably would be the PCIe bus. Is that 92% the BUS monitoring from Rivatunner? I am checking also FB (framebuffer?) but that one is not clear do me what it means. Added. By the way what are you running to reach that BUS % usage?
That pic is from MSI afterburner bus usage > I7 4770k@4,8GHz, 2666MHz ram, 1080Ti > PCIe@2.0x16 with all setting max, 76 AI, weather Storm&Overcast
There's enough PCI discussion already, please show some courtesy to the OP and stay closer to the topic
I do not agree with you... @Suzukinol shortly I will update to i5 8600k with the gtx 1070 in games the graphic strip much more than the cpu the improvement will be remarkable by the way I will have pcie 3.0 , I leave the FX 8350 and switch to Intel, in mononucleo in much better
Same card as me. Ran brilliantly in the rain at Adelaide the other night. Only frame drops were on the TV cams.
Well, I could try today with Silverstone + USF 2000 I am getting 100 FPS on Dry and 54-60 FPS on Wet.... Oh please guys, the rain needs a serious optimization. I don't feel normal to loose 40 fps just because of the rain.. I used these settings: https://i.imgur.com/CSiptVK.jpg https://i.imgur.com/YPM5jn2.png Cheers.
I can also see a noticeable fps drop when you are in the garage menu of about 10 - 30 fps (depends) on Rain and 5 - 15 fps on Dry.
It is known issue that there is huge fps drop when wet weather is used. Road Map February 2018: Performance We mentioned last month already that performance was being worked on, it’s still one of our main priorities, we’re continuing to chip away at it and have come even further, it should not be long now until we can release these improvements to you. It’s safe to say you’ll will be very surprised at the end results. In our internal testing we’ve seen a substantial performance boost and higher frame-rates across the board. As with anything though we are cautious, it’s in the hands of our internal test team to make sure it’s all working as expected.
Well he is using most settings maxed or high. Rain will basically half FPS on every GPU and with a 1050 Ti if you add 20 GT3 cars you don't have that much FPS room to play with.
Yeah, could be. My test was with 20 AI and basically using ultra settings but I was getting 100 FPS on dry so it was like: "mmmm????!!"