Any form of decent Anti-Aliasing would be nice, not this MLAA/FXAA PP bullsh*it
I could say, the entire kit. Which the cpu, gpu and display playing a good factor on it. I mean displays improved a lot these days. If you consider KarQ most of his values at 273fps were higher on display not exactly game. Also his gpu was like 1650 at 73 fps on some portion of his video.
But for sure cpu on the video and display on their claim are the major culprints. Most of the current gen cards handle 1440p well enough, dunno about stable framerate but epic 1440p at 150% my rx 6900xt lc(oem) handles really fine those framerates, while the fan gets loud as heck. I undervolt it btw still runs 220fps+ low noise and stable with about 67-69°C hotspot. While if I run at 400 would be really loud and at about 89°C hotspot. Later I would consider get 360mm rad instead of 120mm for it but so far 1080/1440p this thing never let me down at max settings and I donât care much about ray traycing anyways.
I have 21-23 fps on a8-4500m for gods sake. Isnât a good apu at all. Reflex and most stuff from nvidia are oversold. Doesnât mean they donât improve, just their claims are âexaggeratedâ. The game barely uses 79% of my gpu and 68% of my cpu.
Prior to it I had r9 380, which had lower SIM than my gtx 1080. Around the time they implemented reflex on OW, I tried it out and the value werenât stable but had better lows and overall had less SIM values on OW than my r9 380. Which was funny at similar fps my gtx had way worst SIM values than R9, but with reflex gtx had slightly better but the SIM was all over the place.
I think the cap from the game doesnât actually is a hard cap, if you have v-sync and reduced buffering disabled. I simply noticed due having a second screen and was monitoring their stats trying to make some sense on SIM, also due SIM being sensible to disk, cpu and gpu, so after some experiments I reached that Pc latency would be the âequivalentâ under the âsystem latency metricâ, only leaving behind display and mouse. The radeon software shows frame time which matched with one of 3 SIM values.
The âbypass were showed on radeon softwareâ the game didnât showed the 400fps. But the 2.3 and 1.8 frame time both appeared on the game while on the adrenalin showed over 400 fps. Maybe because I had 2 screens or the game simply donât exactly âstopsâ just âdemotesâ the usage, having ocasionally those âbypassesâ, similarly with the 5-9fps drops on certain moments not being locked at 400 without triple buffering or v-sync.
My advice would be go on practice do some random moves, shots, flicks and jumps checking the fps. Pick the lowest value and add about 5-20 fps. Test how the SIM responds, if stays stable you will have better experience. Until you narrow to a fps that almost stays the same like 95-99% of the time. Always focus on stability than uncapped, your aim and consistency will be way better than oscilate from 400 to 270 by example.
Try to match the speed with your monitor to have the âbestâ case scenario.
Uncapped only actually helps on loading times, but after certain value the benefit from it is meaningless if you donât have stable frames. Which the 4090 didnât, having variance almost as double during the gameplay.