5800X3D is still king.
Is it still better than i9-13900K ?
Yes, Iāve seen a benchmark graph yesterday but I canāt find it now. Will update when I do.
How about performance 3090 pairs with 13900K?
Edited: now, everywhere reports 13900K alone draws as much as 300W
I guess 750W PSU would be enough then.
Total system usage of almost 500W on all-core max load according to hardware unboxed.
Itās a graphics card for $2,000! That kind of money bought you a complete high end gaming PC just four years ago, top of the line graphics card included!
Itās insanity and it better be WOW for that kind of gravy!
Yes I agree. Whatās even more insane is the amount of people with this kind of money to spare for a graphics card. I see so many people upgrading from their 3090 like itās now redundant
At what point is enough fps? If the RTX 4090 is the holy grail then nobody will buy the RTX 5000 series next year. Then everyone will say Iām getting 250fps instead of 150fps. Surely 150fps is enough for flight sim? Iām perfectly happy with 50fps on my RTX 3070.
For me anything more than 35 is almost fluid and I can get that even with my humble i7-10700K abd RTX 3080 despite using FSLTL as well as 200 TLOD,LOD.
Iām rocking a 5600X and 1080Ti, current GPU bottleneck, so Iāll be watching out on the 4070 release to pick up something which is balance for my CPU while not being too expensive.
Nice. What is the rest of your PC? Was it a custom self-build or modified pre-build?
Interesting. Safe to assume that the 13700k would be sitting between the 600 and 900?
Unfortunately it is a CPU limitation
At the same airports with similar settings, my 3090 and i5-12600K at 3440x1440 achieves 40-45 FPS while sitting at the gate in the flightdeck, with the 3090 struggling to reach 55% load.
I ordered an i9-13900K to balance the 3090ās performance.
I think the 4070 will probably be overkill for a 5600x CPU. I would recommend getting a 30xx series GPU on the cheap.
Not sure if itās been mentioned in this thread, but for the best image quality, frame generation can be combined with TAA.
Try to spot image distortions due to AI frame generation. As the GPU generates a frame in between two traditionally CPU-rendered frames, the chance of seeing them should be 50%.
Iād expect the most artifacting at the boundaries of objects moving in different directions/speeds on screen (for instance a fast-moving background against a still part of the cockpit). I think I do see some slight artifacting in several of those images at the edges of the cockpit, but itās hard to tell when thereās so little apparent motion on screen.
Try some scenes where youāre actively panning the camera around, especially if you have a head-tracking system (people with head trackers are really particular, and this is similar to how VR works).
Panning and translating the camera will create fast-moving near parts of the cockpit against slow-moving background objects, while everything is moving around in various directions. Should be a much better test case.
How is it a waste when i cant achieve a solid 30 fps in aerosoft ebbr with fsltl on? And i have a 3080 card.
I think there are several ways to use a RTX4000-series card. Like I mentioned one possibility is TAA + frame generation for the best image quality. Then thereās the different DLSS-options. Iāve tried both DLSS quality (upscaled 1440p for my 4K display) and DLSS performance (upscaled 1080p for my display - that turned out too blurry for me) - so this will depend on the display resolution and whether youāre GPU-limited. Which I donāt think will be a huge issue with RTX4000-series.
I think the key thing is frame generation which will be available for all the RTX4000 cards. For the 4090 at least, this effectively doubles the fps, and I canāt see how it would be different for the 4080 etc.
If you have e.g. 28 fps at EBBR then frame-generation will give you twice that i.e. 56fps.