How is Nvidia Reflex different from the 'Prefer Maximum Performance' Power Management setting?

[Actual_Criminal]

Nvidia Reflex has recently been introduced on Overwatch and I have been watching some reviews on YouTube of it.

 

A user said that it basically forces your GPU to work at full load to make full use of it even in a game where it may only need 40% performance (example) to operate the game at max settings, which therefore creates more frames.

 

How does that differ from the Prefer Maximum Performance option under Power Management in the Nvidia Control Panel? - The description of this sounds pretty much like NR. 

 

Can anyone explain it better? Obviously, there is a difference as the Max perf option has been there for ages.

 

 

[SkilledRebuilds]

Reflex sends images to the queue or from the queue "fast as it can" disregaridng other factors like MAXPERF does.

Relfex does "Fastest..JUST IN TIME" or... "Fastest end of the chain" where MaxPerf just keeps clocks high but doesnt mess with frame queuing/delivery methods.

[Mark Kaine]

9 hours ago, SkilledRebuilds said:

Reflex sends images to the queue or from the queue "fast as it can" disregaridng other factors like MAXPERF does.

Relfex does "Fastest..JUST IN TIME" or... "Fastest end of the chain" where MaxPerf just keeps clocks high but doesnt mess with frame queuing/delivery methods.

soo, what's the difference?  Moar FPS I  hope?!

 

Fun fact: prefer maximum performance cost me actually performance because the card runs hot as hell and downclocks at any chance it gets (about 200 points in Superposition,  and that can only get worse because it's a really short test)

 

 

But yeah, I'm curious,  what does this "Reflex" thing really? 

[Tieox]

I've a Corsair PSU and I disabled "prefer maximum performance" due to similar reasons.

 

1. It constantly kept triggering the fans to activate on the GPU due to rising temps as mentioned.
2. The coil whine from my Corsair PSU was annoying AF.

 

[KaitouX]

3 minutes ago, Mark Kaine said:

But yeah, I'm curious,  what does this "Reflex" thing really? 

Lowers the latency.

[Mark Kaine]

6 minutes ago, Tieox said:

I've a Corsair PSU and I disabled "prefer maximum performance" due to similar reasons.

 

1. It constantly kept triggering the fans to activate on the GPU due to rising temps as mentioned.
2. The coil whine from my Corsair PSU was annoying AF.

 

that's interesting... i have a corsair too... and a 3070....

 

 

The FUNNY thing is I have a 500w bequiet and, the whole PC is running so much cooler with it,  and doesn't crash or anything, it's just cooler...

 

 

I'll probably put it back in summer cause I can't see me enduring this heat and noise...  seriously something with the corsair fan gotta be wrong also  it's loud AF at "800rpm", shouldn't be audible at all really! 

[Tieox]

3 minutes ago, Mark Kaine said:

that's interesting... i have a corsair too... and a 3070....

 

 

The FUNNY thing is I have a 500w bequiet and, the whole PC is running so much cooler with it,  and doesn't crash or anything, it's just cooler...

 

 

I'll probably put it back in summer cause I can't see me enduring this heat and noise...  seriously something with the corsair fan gotta be wrong also  it's loud AF at "800rpm", shouldn't be audible at all really! 

I won't be keeping the Corsair PSU, strongly tempted to take the EVGA 650 G3 from my old system and plonk it in this (swap psu's basically), but also thinking of future expandability so likely will go for a 850 EVGA PSU when funds permit. 

 

This is my second run in with Corsair PSU's and the second time I've ended up removing it for an EVGA , I got a screaming deal on the system, and a turn around in just over a month for these specs? quite suited to have to swap a PSU for coil whine as the only problem!

[SkilledRebuilds]

MaxP works well with Undervolting.

Locking in a MHZ that wont top the PWR under full loads.

0.95v and 1920-1950Mhz vs 1.037v 1860-1920Mhz

Or underutilized (1080p Low esports) where GPU usages are lower, temps lower but you MaxP for Latency (running MAXP 1980Mhz flat at 60% usage at 1080p vs NotMAXP and dropping clocks to 1200-1400 or less, switching a lot and adding increased latency to esports gaming)

 

So you shouldn't use MAXP when at 90-100% GPU Usage by those and your experiences in the replies here.

Only when sub80% GPU then activate MAXP in competative environments.

 

 

Don't know of any Reflex negatives because I've not specifically looked it up.