• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

NVIDIA Reflex Tested with LDAT v2

Inle

Staff member
Joined
Apr 6, 2017
Messages
341 (0.12/day)
System Name Efrafa
Processor Intel Core i7-5960X @ 4,3 GHz
Motherboard Asus X99 STRIX Gaming
Cooling NZXT Kraken X52
Memory G.Skill Ripjaws 4 32 GB
Video Card(s) Asus ROG STRIX GeForce GTX 1080 OC Edition
Storage ADATA SX8000 NVMe 512 GB + 5x Kingston HyperX Savage 512 GB
Display(s) Acer Predator XB271HU
Case Corsair Crystal 460X
Audio Device(s) Audiolab M-DAC
Power Supply Seasonic X-850
Mouse Logitech G900 Chaos Spectrum
Keyboard Cherry MX Board 6.0
Software Battlefield 1
NVIDIA Reflex is a collection of advanced technologies set out to drastically reduce end-to-end system latency and make you a bigger threat on the virtual battlefield. We used the latest iteration of the Latency Display Analysis Tool to thoroughly test how Reflex behaves in practice.

Show full review
 
Battlenonsense also tested this. In essence, Nvidia reflex does what limiting your GPU utilization to 98% does with an in-game frame limiter. Of course it only has an impact when GPU utilization is maxed out as that's when you see an large increase in latency.
 
This is the type of content TPU deserves

This also explains why some people have very different views on Vsync causing input lag, when others have no issues with it - all comes down to the hardware
 
Last edited:
Nice to know that Reflex also help a little in CPU bound scenario, some games like Warzone would run into cpu bottlenecking especially with DLSS enabled.
 
This is the type of content TPU deserves

This also explains why some people have very different views on Vsync causing input lag, when others have no issues with it - all comes down to the hardware
Ya Vsync always a hot point for discussion. I think sometimes it depends on the game or game engine also. My screen is 60Hz, I find vsync helps lots for quite a few games for tearing, as for any additional input lag it's trivial compared to the amount of tearing, or lack there of after vsync is on.

Also I don't think anything could make me a better gamer lol playing 2 hours a month isnt gonna do it.
 
Last edited:
"LDAT ... might strike you as something that was put together as a school project."

I must have gone to the wrong school!
 
Now if only I can get my hands on a GPU, I would love to test this myself, I am extremely sensitive to input latency.
 
I suspect we will see more nice things from nvidia to "battle" the vsync, latency and frame pacing issues breaking the immersion.
Think about it, with their deep learning AI uses for reconstructing frame data for raytracing and DLSS, they'll soon start guessing incomplete frames to never miss a pace. I can even see the banner on driver installer: constant low latency with constant high framerate.
 
Enabled it and can't tell any difference in Overwatch.
Does that mean I just plain and simply suck?
 
Enabled it and can't tell any difference in Overwatch.
Does that mean I just plain and simply suck?
or you were already running low latency
 
Seems nice, but there's definitely nothing that can help me with my shitty aim lol.
 
Seems nice, but there's definitely nothing that can help me with my shitty aim lol.
:laugh:you have to wait than til nvidia come out with AAKA v.1 (auto aim kill all) :laugh:
 
"What I can tell you from firsthand experience is that activating the Reflex technology in a GPU-bound scenario instantly makes the game feel significantly more responsive. Mouse movement feels snappier and more direct, which translates into a vastly improved overall experience."

Honestly, your impression is worth absolutely nothing in this case, because you're not doing a blind test. Your brain (as well as that of any human being) is highly prone to confirmation bias. To test this, you should tell someone you know to randomly (preferably using a coin throw or random number generator) turn Reflex on or off for you, but not tell you which setting it's on, and then you note down whether you think it's on or off based on how "snappy" the game is. Ideally this person should not know what the test is about so as to not influence you through subconscious body language. After 16 repetitions of this test and comparing your notes versus that person's (which contain the setting it was on), it should be enough to verify whether you REALLY can tell the difference or your brain is just fooling you.
 
Last edited:
"What I can tell you from firsthand experience is that activating the Reflex technology in a GPU-bound scenario instantly makes the game feel significantly more responsive. Mouse movement feels snappier and more direct, which translates into a vastly improved overall experience."

Honestly, your impression is worth absolutely nothing in this case, because you're not doing a blind test. Your brain (as that of any human being) is highly prone to confirmation bias. To test this, you should tell someone you know to randomly (preferably using a coin throw or random number generator) turn Reflex on or off for you, but not tell you which setting it's on, and then you note down whether you think it's on or off based on how "snappy" the game is. It's also important to ensure this person leaves the room after each time they switch (or don't switch) the setting to avoid them unconsciously influencing you through body language. After about 16 repetitions of this test it should be enough to verify whether you REALLY can tell the difference or your brain is just fooling you.
yeah if he has the testing and sets things up for the highest latency result and then does a simple on/off toggle, he's gunna give a real opinion of whether it helps or not
 
someone you know to randomly
I disagree, you need a gamer audience, not randoms off the street. We care about improvements for us, not for my grandma
 
  • Like
Reactions: las
I disagree, you need a gamer audience, not randoms off the street. We care about improvements for us, not for my grandma

Err, the grandma won't be emitting any opinions on what she sees, she's only pressing a button to switch Reflex on or off and the article author will do the testing. Surely you don't need to be a gamer to click a button properly?
 
1621782894478.png


I didn't understand this portion of the results from LDAT. From what you said in the article, Reflex basically prevents GPU overutilization while Boost raises GPU clocks for CPU bound scenarios. So how come you got benefits from non-boost Reflex when nerfing your CPU? Shouldn't what is essentially a dynamic frame limiter be useless when you're CPU limited?
 
Back
Top