Summary
In the game Valorant we could see that the two tested graphics cards are basically on par. The only exception was in UHD resolution, here NVIDIA Reflex could still lower the latency by about 3 ms. It remains to be seen to what extent this is decisive in the end. CoD MW2 once again proved that Radeon Anti-Lag doesn’t really work. With Anti-Lag deactivated, the latencies are one millisecond better than with Anti-Lag activated. In CoD MW2, only those who can use Reflex benefit in the end.
In Overwatch 2, we could actually see both technologies in use. Radeon Anti-Lag, when it works, can reduce latencies. At the end of the day, however, Anti-Lag takes a significant beating. NVIDIA Reflex pulls up and away. I’ve also made two nice slides for you that show: That you can play with an RTX 3070 Ti in UHD at around 100 FPS and have better latencies than with an RX 6700 XT in 1080p at over 200 FPS. Fact: NVIDIA Reflex wins – flawless victory!
AMD will have to come up with something to keep up with NVIDIA. Frames win games is unfortunately only half the truth. First and foremost, the FPS and 1% low determine how smooth a game feels – depending on the monitor (refresh rate). Because 200 FPS on a 60 Hz panel are rather meager in the eSports area! The latency in the game ultimately decides how responsive you can hit a target. I was able to try this out live in a self-test. If you combine high FPS with low latency on a 240 Hz/360 Hz panel, you really notice a difference.
Conclusion
When anti-lag works, it can reduce latencies. In Overwatch 2, that was the case. Valorant still runs in the CPU limit even in UHD, where the Intel Core I9-12900K is weaker. I cannot prove whether a 13900K would do better due to the lack of a CPU. The fact is, NVIDIA Reflex is and remains the number one for the time being when it comes to latencies. NVIDIA’s approach – from within the game, so to speak – to clear the congestion in the render pipeline is not only more efficient, but much more effective. Radeon Anti-Lag continues to take the driver-based approach that NVIDIA used to take with ZERO (NVIDIA Ultra Low Latency). You can do it that way – but you don’t have to!
FYI: For DX11 based games, you can still enable NULL via the NVIDIA control panel. I always have NULL enabled because it doesn’t bring any disadvantages. The effect of NULL is in the end comparable to Anti-Lag. Of course, you take that with you when Reflex is not available.
Outlook:
I’ll stay on the topic of Anti-Lag, I won’t let AMD off the hook so easily. Furthermore, I’ll take another look at the three games in 1440p on the PG279QM. Then I’ll measure the actual end-to-end latency with LDAT and compare the values from FrameView with those from LDAT. This should be interesting.
78 Antworten
Kommentar
Lade neue Kommentare
Urgestein
Urgestein
Mitglied
Moderator
Moderator
Veteran
Veteran
Veteran
Moderator
Urgestein
Urgestein
Urgestein
Urgestein
Mitglied
Urgestein
Urgestein
Urgestein
Urgestein
Alle Kommentare lesen unter igor´sLAB Community →