NVIDIA GeForce Driver Power Mode Settings Compared

Thanks for doing this testing. It's something I've always wondered about as I've tried tweaking many of the settings in CP over the years. Back when I ran SLI setups I saw similar unusual metrics that you came across but additionally on some rigs having it set to maximum performance caused instability issues even when I removed any other custom settings or over clocks.

Just curious. What are your experiences with the texture filtering settings? I've some unusual fps effects with different games. Some seemed to gain with highest while others definitely lost and were better at the default.

I'm happy that Nvidia has continued to add features to control panel over the years but unfortunately their explanation and effect on gaming seems to have diminished over time as well.

Thank you again for this in depth testing!
 
Wow, that was an impressive amount of testing. An underwhelming result, but the review is very much appreciated. Thank you for diving so deeply into this for the masses!
 
There are some specific games which do not play well if the Optimal Power setting is used; at one point PubG ran like garbage if you had Optimal Power selected and going to adaptive fixed the issue. I've seen others, but it's been a little while so I don't recall what they were. Good job on this!
 
Interesting.

I always just popped it in "prefer maximum performance" and left it there, since I only ever boot into widnows to play games, so I didn't think the higher idle power was a problem.

I had considered that this may hit thermal limits and prevent the GPU from boosting as high as it otherwise would, but I thought the chances of this were slim, as I have a water block on it and never go above 38-40C at load.

I had never considered that the power limit might be a culprit and limit performance.

Based on this I may drop it back to optimal. Either way the difference is tiny though.
 
lostin3d said:
Thanks for doing this testing. It's something I've always wondered about as I've tried tweaking many of the settings in CP over the years. Back when I ran SLI setups I saw similar unusual metrics that you came across but additionally on some rigs having it set to maximum performance caused instability issues even when I removed any other custom settings or over clocks.

Just curious. What are your experiences with the texture filtering settings? I've some unusual fps effects with different games. Some seemed to gain with highest while others definitely lost and were better at the default.

I'm happy that Nvidia has continued to add features to control panel over the years but unfortunately their explanation and effect on gaming seems to have diminished over time as well.

Thank you again for this in depth testing!
Click to expand...

Now, on the texture filtering settings, that's a whole other ballpark to test. It's a completely separate issue, of course, not related to power, but could be related to performance. I don't really have an answer for you right now. I haven't done an AB comparison, I haven't looked at performance in regards to its options.

I can tell you this though, on my personal gaming system (granted I have an RTX 2080 Ti) I put it on the highest quality. However, I have a feeling it might affect performance to a noticeable degree on lower-end to mid-range cards. I'm a stickler for image quality, I like good image quality, so I do this to ensure the best in my gaming.

However, for review purposes, it makes sense to use the default standard driver settings on both AMD and NVIDIA. This default setting is the gameplay experience both want to offer, so we look at it from that standpoint in reviews. In regards to that, gone are the days of AMD and NVIDIA graphics looking different.

Perhaps in the future, I will fully test the filtering question.
 
Captindecisive said:
There are some specific games which do not play well if the Optimal Power setting is used; at one point PubG ran like garbage if you had Optimal Power selected and going to adaptive fixed the issue. I've seen others, but it's been a little while so I don't recall what they were. Good job on this!
Click to expand...

Noted

Naturally, if you personally find that one setting is advantageous over another, use it.

There are many scenarios, many games, many different system configs. It is hard to predict and test them all. Your case may be different, your mileage may vary. Every game behaves differently.

On the whole, "Optimal Power" is the best setting, if you don't think it is for you, make sure you aren't losing performance or clock speed while gaming in another setting.
 
Zarathustra said:
Interesting.

I always just popped it in "prefer maximum performance" and left it there, since I only ever boot into widnows to play games, so I didn't think the higher idle power was a problem.

I had considered that this may hit thermal limits and prevent the GPU from boosting as high as it otherwise would, but I thought the chances of this were slim, as I have a water block on it and never go above 38-40C at load.

I had never considered that the power limit might be a culprit and limit performance.

Based on this I may drop it back to optimal. Either way the difference is tiny though.
Click to expand...

It certainly warrants testing. Look to see if your clock speed is dropping, and or performance. It is possible as I found out, due to power limits, not heat limits.

Generally speaking "Optimal Power" is the best setting.
 
Very nice topic.
I had always thought of these settings to improve stability rather than FPS.
It would be interesting to see if the "maximum performance" would allow a bigger OC vs optimal or default.
 
Brent_Justice said:
I can tell you this though, on my personal gaming system (granted I have an RTX 2080 Ti) I put it on the highest quality. However, I have a feeling it might affect performance to a noticeable degree on lower-end to mid-range cards. I'm a stickler for image quality, I like good image quality, so I do this to ensure the best in my gaming.
Click to expand...

I do pretty much the same with mine. It's why I sank so much money into it and the plethora of displays I go thru.
 
Good review. Other than idle power, not much was expected to change once the card is in a steady load state. What would be more interesting is if you could test the latency/subjective feel in workloads that are constantly changing between idle and load states. I have found maximum performance mode beneficial while manipulating dense point clouds. Stop, rotate, stop, rotate felt smoother. It would be interesting if you could confirm if it's just placebo or there is something there.
 
Nice test. Don’t think I’ve ever seen a reviewer dive into this before. Also seems to be highly situasional. My 1080ti stabilises at 47C during load, since I have a custom watercooling setup. Therefore I don’t see the same clock frequency discrepancy for running the maximum performance profile.

Meeho brings up an excellent question as well. I would love to see some thoughts on that to.
 
You must log in or register to reply here.
Become a Patron!
Back
Top