GPU vs. CPU limit regarding the system latency
Let’s compare the measured values in terms of latency at different average FPS. To that end, here’s another slide from the last article.
Now who’s going to notice something when you look at the last few slides? Difficult to see! I’ll break it down, check out the next slide! At that point, we discovered an anomaly in the Matrix! What was the claim? The higher the FPS, the lower the latency! Let me think for a second.. No, I don’t believe it! At least not anymore without asking about the exact circumstances. Who would have believed that 165 FPS would have lower latency than 236 FPS? Be honest with yourselves – I think very few!
Why is that now? Simply explained: In the CPU limit the GPU has to wait, so you reach a point where nothing more can be done regarding the latency reduction. Because the CPU can’t keep up with the GPU! At GPU limit, the render pipeline fills up and the CPU pushes in more and more. Ergo: Latency goes up! Here, the GPU can no longer keep up with the CPU. If you reach this GPU limit (99% GPU utilization), then either NVIDIA Reflex, the reduction of the resolution or the reduction of the settings or a manually set frame limit will help. The frame limiter always works, no matter if AMD or NVIDIA GPU. Maybe Radeon Boost and Anti-Lag will help too? Let’s see what the measurements reveal over the next few days!
So I think at this point it should be enough to understand. Next or last topic and also last page for today!
30 Antworten
Kommentar
Lade neue Kommentare
Veteran
Urgestein
Veteran
Urgestein
Urgestein
Veteran
Urgestein
Urgestein
Veteran
Urgestein
Urgestein
Urgestein
Urgestein
Urgestein
Urgestein
Urgestein
Veteran
Alle Kommentare lesen unter igor´sLAB Community →