Nvidia Optimus reduces gaming performance

In 3DMark or in games? Keep in mind that 3DMark is GPU-bound and runs at very low frame rates, hence is not likely to brush up against this limit.

 

Meanwhile I have Optimus reducing battery life:

 

Definitely a MUX chip by that wording. I'm surprised they don't even acknowledge it as a feature or allow G-Sync to function.

It costs them money and it's a desirable feature for Gamers. ASUS made a huge deal about it in their rigs when they adopted it.

 

Same thing with my Acer Triton 500, it has the option to switch GPU’s in the bios.
I actively crossed off a lot of laptops off my list because of no Gsync.

 

great thread, thanks for info.

 

Smart thinking. Gsync panels have the extra advantage besides better FPS.

 

wow i didnt know that G-Sync or no G-Sync would have that much better games performance. Next time I am in the market for a new laptop I will stick to G-Sync only laptops.

 

It’s not G-Sync that has better performance, it’s the lack of Optimus.

 

oh yes, i see that now

 

Precisely why I am looking at the Asus models. They have a toggle to disable optimus and use only the dGPU.

 

I have a MSI GS65 w/ 2060RTX, this notebooks have Optimus and not GSYNC. But what happen if I use a external monitor via Mini Display Port? Can I use GSYNC in a compatible monitor?

I will have the performance boost due to iGPU are off? With the integrated display iGPU are always ON, but with external monitor?

 

I also reported different GPU performance on my M6700 when I was running dual internal lcd, one was a LVDS 1080p panel controlled by the intel HD4000 and the other one was an eDP 4k panel controlled by the P4000, so it's basically one an hybrid setup where one of the lcd is affected by "optimus", I didn't measure the performance gap but it was clear in Unigine superposition benchmark

 

I made some benchmarks, internal display (optimus) vs external display (nvidia gpu):

https://www.reddit.com/r/laptops/comments/dkt0rv/nvidia_optimus_vs_external_monitor_huge_fps/

 

Performance is almost exactly the same on the Precision 7740 with Quadro RTX 4000 Optimus vs. Dedicated:

https://www.3dmark.com/compare/spy/8967012/spy/8966960

Definitely get a second stick of ram. The iGPU uses system ram for the framebuffer so it likely limits high framerates unless you have dual channel.

 

Second everyone else's suggestion about getting another stick of RAM for dual channel and doing the comparison again.

Synthetic benchmarks like Time Spy already run at very low frame rates (ergo low PCIe bus utilization) and low CPU utilization, so they don't show Optimus bottlenecks like real games running at high frame rates do.

 

I just don't know if this is universal. For example in all of the testing I can find the GT76 performs on the level with the Area 51M and the high-end desktop CPU Clevo machines.

 

Do you happen to know what those tests were?

 

I've seen a 9980HK ROG Mothership outperform a 9900K A51m.

 

As it should, since it has better cooling and can support memory faster than 2400MHz. It supports both G-Sync/discrete-only and Optimus via a mux switch.