I have, what may be a dumb/basic question about the GPU:
My old laptop had a Nvidea 1060 and I saw that MSFS would run it at 97-99% utilization almost constantly.
I just started using a new laptop with a 3070 and it too is running at 97-99% using the same MSFS data settings.
Is that normal? I had expected the new one to run much less.
depends on your settings. Are you running both at the same settings or higher settings on 3070? I run 3080ti and it runs 100% utilization in the menu. lol.
Yes, the GPU is designed to run flat out at 100% without issue. It’s the CPU that starts to run out of steam after 50% and is basically swamped after 85%.
I just upgraded to a 3090. According to the Xbox Gamebar, it’s running at 97% GPU usage. Only something like 17% VRAM usage (3090 has 24 Gb), something like 33% CPU usage (average of all cores? mine is an i9-9900K); this is in 2D. So what is the GPU doing at 97% usage? It’s giving me ~60 fps over Manhattan in ~all ULTRA settings at ~1500 ft, 100 kts, vs. 30 fps with a 2070S. So being an ignorant newbie, it would seem that the GPU in SU5 can now get cranked up to the max to take full advantage of what you’ve got in your computer. Although it would seem like I really don’t need what I’ve got with the 3090 and it’s overkill for the VRAM, etc., I haven’t run it with my Reverb G2 yet. Here’s hoping I can run in close to all ULTRA settings there, too, and I expect that I’ll be taxing it a lot more in VR!
Thanks everyone. It sounds like it’s normal and the new 3070 is just providing a better MSFS graphics. I need to go to fly some places I’ve been before to see a comparison.
Might be my ignorance of MSI Afterburner but it only seems to show me “GPU temperature.” If I run CPUID’s free HWMonitor app or TechPowerUp’s free GPU-Z app, those apps show in addition the GPU memory and the GPU “hotspot” temperatures. The worst is GPU memory, at least for the 3090 and it’s already been mentioned on the forum. In spite of reaching a memory temperature of 94 deg C (when the supposed thermal limit of the card is 93 deg C), my 3090 FE card fan was only running at about 1,100 rpm. I used MSI Afterburner to crank both fan speeds up to 2,300 rpm, which according to the MSI Afterburner curve is close to max fan speed (a review says max is 2,600 rpm). The GPU temperature fell to 53 deg C and the GPU memory to < 80 deg C. It’s a bit noisy but fits right in with the GA aircraft and helicopters I like to fly! So although in terms of sim performance it’s good to max out your GPU usage, if you have a plain vanilla FE 3090 as I do or something similar, you might want to carefully check your GPU memory and hotspot temperatures with one of the apps that I mention. My room temperature is around 28 deg C (83 deg F). The 3090 is sitting outside my computer all by itself on a riser stand so it should be getting the max possible ventilation.
97%+ GPU usage means you’re playing at high graphics detail and/or high display/render resolution. Although I wouldn’t rule out the Xbox game bar causing performance problems and/or reporting erroneous data…
GPU utilization is usually a function of resolution. Anything below 4K is typically bottlenecked at the processor, so it’s completely normal to see low utilization (~50%) of a GPU at 1080p or 1440p resolutions. At 4K, usually the bottleneck is the GPU and you’ll see close to 100% utilization.