Page 2 - Question - Are These Temperatures Too Hot for MSI RTX 3090ti Suprim X? (2024)

Hi, I recently bought an RTX 3090ti card and was wondering if the following gpu temps is something i should be concerned of. Ive been sitting on this thought for a little bit and wanted to get it off my chest. When gpu temps increase so does cpu temps. My cpu is an i7 8700k cooled with an 120mm corsair h55 AIO installed at the rear side of chassis as exhaust, so the hot air produced inside the chassis by the gpu fan goes through the radiator out the case. My case is a white Be Quiet! Pure Base 500 with tempered window and installed with max chassis fan which has two 140mm exhaust at top and 3 120mm fans at from as cool intake all of which are Be Quiet! brand fans.

The following from GPU-Z app while running microsoft flight simulator 2020 with cpu temps rising from mid to high 70s degrees based on the figures below:
GPU Clock: 1995.0Mhz
Memory clock: 1312.7MHz
GPU Temperature: 77.3 Degrees
Hot Spot: 88.1 Degrees
Memory Temperature: 72 Degrees
Fan 1 Speed (%): 75%
Fan 1 Speed (RPM): 2432 RPM
Fan 2 Speed (%): 75%
Fan 2 Speed (RPM): 2436 RPM
Memory used: 8730MB
GPU load: 94%
Memory controller load: 24 %
Video Engine Load: 0 %
Bus interface Load: 5 %
Board Power draw: 445.9W
GPU Chip Power Draw: 357.9W

Jabad96 said:

what i dont understand is which would cause thermal throttling, since there is the gpu, hotspot, and memory temps.

Click to expand...

They all can. It's just that their conditions are different.
Gpu temperature, which is an average, is good up to 82.5C, or 83C for the 3090Ti. It has a maximum of 90C. This is one of the parameters Gpu Boost bases the gpu's boost bins on.
Hotspot is the hottest sensor, of which there are many, on the gpu die, and is good up to 110C.
Memory is good up to 105C. There is a general recommendation out there for 95C or lower for SUSTAINED loads - spikes don't count.

Jabad96 said:

My cpu temperatures have dropped after putting the radiator at the front. Now it sits around 50s to low 60s when running the games. Then I decided to turn on intel turbo boost which is 4.7ghz and cpu temp topped at 81c for on core for flight simulator 2020, and saints row 2022. I did this to try to get the most out of my cpu with out manual overclocking which i dont have experience with. I realized im in need of a new cpu cooler since i had the corsair h55 120mm for about 9 years. I am deciding to post pone purchasing a new one since ive decided to get a new cpu raptor lake when it comes out including motherboard and ddr5 ram because my current setup with corsair 16gb 3200mhz red led and i7 8700k is quite a bottleneck for this gpu. So im currently ok with higher cpu temps with turbo boost under gaming load. Ill probably end up getting a corsair aio. Ive tried installing deepcool ls 720 360mm but the heatsink wouldnt completely sit on my cpu regardless of how hard i screwed it and dont like the idea of applying that much pressure on mobo so i returned it after following all their installation instructions for lga 1151. Ive also decided to swap the msi suprim x with asus rog strix lc oc edition gpu and now have additional space in chassis slightly high clock speed out of box with gaming and cooler temps. I mounted the radiator for the gpu up top 240mm since i have the cpu aio 120mm mounted at front. I am getting around mid 50s to low 60s under gpu load but what i dont understand is which would cause thermal throttling, since there is the gpu, hotspot, and memory temps. The memory temp and hot spot temp are usually identical, is this how it is suppose to be?

Click to expand...

The memory temp probably IS the "hotspot" temp, since the hotspot can be whatever sensor is the hottest.

Is your AIO for the graphics card that is mounted at the top blowing out or pulling air in? Same question for the AIO you moved to the front? And did you put a fan in the rear location where the AIO was previously located, as an exhaust?