Question87 Celcius Hotspot and 2500 rpm fan speed on a stock brand new GPU ?
Question87 Celcius Hotspot and 2500 rpm fan speed on a stock brand new GPU ?
I bought an ASUS PRIME 9070 XT OC yesterday and didn't understand the fan noise issue despite having a GPU temperature of 58°C. I checked the AMD panel, which showed a hotspot of 87°C and a fan speed around 2550 RPM on the default settings. Is this typical for a brand new card, or are 9070 XTs just running hotter? I also re-applied Arctic MX6 thermal paste, but I'm unsure if that affected the warranty. Many 9070 XTs with hotspots above 85°C run between 1500-1700 RPM, which seems unusually high for this model. My temps are consistently higher than expected.
When sharing a troubleshooting thread, it's important to provide complete system details. List the specifications as follows:
CPU:
CPU cooler:
Motherboard:
Ram:
SSD/HDD:
GPU:
PSU:
Chassis:
OS:
Monitor:
Include the age of the PSU along with its make and model. Also mention the current BIOS version for your motherboard.
You mentioned the GPU is brand new. What GPU were you using before upgrading to the Asus Prime RX 9070 XT? Have you used DDU before installing the new GPU?
It's recommended to use DDU in Safe Mode to clear all GPU drivers (Intel, Nvidia, AMD), restart, and then manually install the latest driver from AMD's support site via an elevated command—such as right-clicking the installer and selecting Run as Administrator.
I also re-applied an Arctic MX6 thermal paste, but I'm unsure if it voided the warranty.
You altered the sticker on the back of your card where the screws are, which likely voids the warranty.
What are your current ambient air temperatures?
CPU model: 13400f
Thermal paste: cooler master hyper 212
Boot device: msi h610m-e ddr4
Memory: 2x8 3200 mhz
Storage: samsung 1 tb ssd (exact model uncertain)
Graphics card: asus prime 9070XT oc edition
Power supply: cooler master 750W Gold 80+ v2
Case: xpg defender pro
Operating system: W10 Home
Display: asus tuf 27' gaming monitor
Previously used sapphire pulse 9070 with a zalman z11 case; at that time the GPU clearance was 290 mm, so I had to buy a smaller 9070. After changing the case and selling the old one, I installed the new xpg defender and started using the 9070XT. I didn’t use DDU or format the PC for a fresh install. The room temperature is around 18-20°C. One thing to note is that the new xpg defender case includes two front intake fans—top and mid—but no fan at the bottom, which might influence GPU cooling. Despite this, if airflow is insufficient, I still see good GPU temperatures (around 30°C). My temperature differences are about 58-88°C.
Image: https://imgur.com/3iPqYhb
Alternative model:
Idle temps – GPU: 39°C, hotspot: 41°C
Game launch – hotspot jumps to 87°C in seconds, fans spin wildly
Game pause – hotspot drops to 50-51°C in seconds
Observations: Such rapid fluctuations seem unusual for me; I’m not an expert.
heat is being trapped inside your case. a solution would be using an AIO, so instead of heat escaping through the CPU inside the case, it would spread through the AIO radiator located at the top and exit. also, ensure your fans are directing airflow correctly—front to back.
It looks like the cooler or TIM on the card might be the problem, even though your GPU temps are normal and the hotspot delta is high. It's not significantly worse than the poorest 9070 XT models tested by TPU, but the fan speed is quite high. If this were an older model, I'd recommend taking it apart to inspect thermal pads and possibly repaste, though with a new card I wouldn't bother.
I disassembled it already and used the new thermal paste Arctic MX6, but it didn’t help. I need to run it at low PL and UV, which means I can’t even manually overclock it. The worst part is I might not be able to sell it when AMD releases a new card, so there’s definitely an issue with this board.
Never used DDU before. That’s your first task. You didn’t mention the age of your PSU or the BIOS version you’re running. If your temperatures didn’t drop after reapplying thermal paste, it could mean there’s not enough cooler air in your system. It might help to test the airflow in your case by removing the front fascia and dust filter to see if temperatures improve. If I were you, I’d consider undervolting the CPU instead of disassembling it.
The main issue is that selling this card at a good price would be difficult because the warranty sticker seems altered. A better solution would be using an AIO system so heat doesn’t build up inside the case.
How did you reach this conclusion? If the airflow in your case isn’t strong enough, a 360mm AIO would struggle, and its cooling power depends on how much cooler air reaches the heatsink or radiator.
If the GPU surface remains intact, it suggests the cooler isn't making proper contact. This is why I avoid disassembling new hardware; I prefer a seamless replacement process. The alternatives I've considered (besides RMA) include using softer thermal pads or swapping them with putty, and possibly adjusting the screws around the GPU slightly.
Based on what thestryker mentioned earlier from TPU, it seems the usual temperature range is present. If your card maintains similar temperatures yet requires rev fans to run more quickly, it suggests a problem with airflow.
Case links:
Front fans: [link]
XPG stacked 2 layers of air filters: one integrated into the front panel and another as an EZ access. This adds complexity and strains the fans, which aren't designed for tight air restrictions.
Check fan RPM with both EZ filter and front panel removed, then repeat with only the front panel in place.
This seems more about minor details; I believe XPG should have considered a different front panel design or kept the large square holes to improve air intake. Currently, the intake appears strongest in the middle, forming a sort of ''.
While the TUF model uses thermal paste, the Prime model used a phase-change thermal pad...
It doesn't seem like the type of TIM (paste/pad) difference matters much, but you may have lost warranty (assuming you're not in the US, where removing the mess would void the warranty if it's illegal).
I receive 85C hotspot around 2300 rpm when I disable the filters at stock settings and adjust -15 PL and -70 mv it drops to 80C with 1700rpm. My PSU is fairly new, purchased three years ago. Should I install another intake fan for the bottom? It seems there might be an airflow issue since the side glass stays warm. Also, changing the default pads could help improve temperatures.
Edit: Removing both filters doesn’t seem to make a big difference; only taking out the front panel appears effective. Removing the EZ filter didn’t improve the temps.