Which matters more depends on your priorities: sharper images or better visual effects.
Which matters more depends on your priorities: sharper images or better visual effects.
I experimented with the benchmark for Black Myth: Wukong recently. I was checking which configurations would let me achieve 4K at 60 FPS on my 4080. I found two good options: one with all settings turned up, including full ray tracing and DLSS at 60%, and another with everything maxed out but medium ray tracing and DLSS at 80% quality. Both had DLSS frame generation active without noticeable issues. In the moving scenes, I didn’t notice a clear visual difference. I wondered which approach people prefer—pushing settings for performance or dialing back for quality and DLSS efficiency. I appreciate how the game lets you adjust resolution sliders from 25% to 100%, letting you pick a preset that matches your GPU’s capabilities and then tweak it to get the best FPS. This level of control over resolution scaling isn’t common, and it’s a fresh way to balance quality and performance. Previously, I’d usually go all-in on settings or lower them gradually, but now I’m enjoying this more precise method.
It varies by game. For titles requiring attention to detail—such as shooters, fake geometry, or low distance clarity—I’ll need to adjust graphics settings to maintain 100% resolution. A more relaxed experience is preferred, like Witcher 3 or CP77. Of course, I’ll lower the resolution scale and enable upscaling for a smoother look.
Results on a 2160p screen have improved significantly since DLSS version 3.7, which is why I now rely mainly on Performance. However, with a 3080 Ti, if performance were even better, I’d likely opt for higher settings and enable DLSS Performance, though I usually turn off Raytracing to keep things manageable for that resolution.
Absolutely right, especially with the RT too. Yeah, I’ve always liked those 32:9 versions but never thought a 4K release would come. I remember some folks doubted these could ever be produced, but once Samsung announced it, I knew I had to save up—no matter how my gear got affected.
Can you operate it using DP 1.4 or HDMI 2.1? It seems DP 2.0 is required for full performance. You might be among the first 5090 buyers.
I tried it recently when it dropped. My setup includes a 4K display, 7800X3D, and 4070. The benchmark results on my system were: Very High, DLSS at 50%, strong motion blur, focus enabled, RT disabled, resulting in about 67 fps on average. Here are some notes. The changes were sequential, so earlier adjustments influenced later ones. Motion blur appeared to worsen disocclusion artifacts, and there was a slight performance boost when I turned it off (average increased from 67 to 69). I didn’t see any blur myself. Disabling FG improved my rate by 44% (from 48 to 69 fps). Leaving RT off had minimal effect, but I noticed poor water reflections and strange glowing sparkles in grass later. This is concerning. Enabling RT Medium caused a noticeable drop (from 48 to 45) and made reflections terrible, plus I saw odd sparkly artifacts. This is unacceptable. Turning on RT Very High was a big hit (from 48 to 34), but it didn’t match the performance I experienced at Medium. Without testing FG, I assume it wouldn’t reach 60 fps. The benchmark showed VRAM usage as low as 7.3GB (no blur, no FG, no RT) and up to 8.8GB (no blur, no FG, RT Very High). I haven’t compared side by side with RT off and RT Very High yet, but from the runs I did, it seems the implementation didn’t make much of a difference. While I generally support RT, this doesn’t look like the best choice for me. RT medium isn’t worth it.
I noticed the distant rasterized shadows had low resolution and some flickering, which is why I initially turned on RT. RT shadows appear smoother and don’t flicker as much. I see the reflections in the muddy creek looked poor with RT on, but they improved noticeably when DLSS scaling increased. The VRAM consumption stayed under 10GB, even at full 4K with DLSS at 100% and FG enabled while RT was maxed out. Overall, this game seems very well optimized in terms of memory usage. Honestly, most of the visuals don’t match the hype around new features; it looks decent but isn’t truly next-gen compared to Cyberpunk, which was released about four years ago on an older engine. Since I didn’t see any visual issues with FG and couldn’t compare input lag between on/off states, I left it enabled. Without FG, I’d likely have to turn off ray tracing entirely to hit 60 fps, even with DLSS around 70-80%.
Appears consistent with my own observations. Medium RT offered little effect, while High RT provided significant improvement, though neither met my expectations. I’m using 1440p with a 5800X3D and 4070, and my defaults matched yours except DLSS was at 75 due to resolution differences. In that setup, I achieved an average of 88 frames without RT. It’s unclear if the hardware unboxing video mentioned Ryzen gains when using the admin account on Windows, but I found this benchmark useful for comparison after running several tests. Interestingly, the software didn’t detect my PC as compatible with Framegen (DLSS or FSR) even when logged in admin. Everything functioned normally otherwise, though I couldn’t activate the setting. I won’t rely on this account for daily use, just wanted to explore it. I haven’t tested other games in that profile or run comparisons without Framegen, but I plan to do so eventually.