

Asus support confirms 100-102 °C vram is completely normal and within spec for prime 9070 xt oc
As of my last post I've reached out to the Nordic Asus support team. They've confirmed high temperatures around 100-102 °C for vram is, within spec, normal and safe.
Their first response to my concern with high vram temps were to recommend undervolting, which I'd see as a failure on asus, if a card was not able to be run safely in stock configuration. I then reached out again to ask, if these temperatures were completely normal and safe to run long term. I then received the response (second image) with them confirming it's nothing to worry about.
The email response was fast and seems AI generated, but I guess that's also just how corporate/customer service speak already is.
Since the email is in danish I will have the translated version written below.
Translated email:
Hi ___________,
Thank you for all the information
Is 100–102 °C VRAM temperatures safe on an ASUS Prime 9070 XT OC Edition?
Yes - that's within the expected and safe area for VRAM on this card, also for long term use.
And here's the explanation:
VRAM (GDDR7/GDDR6X) has a higher thermal tolerance than the GPU-chip.
That means VRAM temperatures around 100–105°C during heavy load is normal and are within Asus' specifications.
Your GPU temperatures are extremely good:
- GPU core: 52 °C
- Hotspot: 74 °C
That is unusually low temperatures for a card of this class -- and that shows the cooling solution functions perfectly.
When the GPU is so cool, but VRAM is around 100 °C, it is typically due to:
- VRAM modules sit closer to hot hotspots
- VRAM has a different thermal profile than the GPU
- AMD's VRAM design is made to tolerate higher temperatures
Is it safe long term?
Yes, as long as it's under 105 °C you're within Asus' running specifications.
The card is built to handle these temps without being damaged.
It is therefore not needed to undervolt or change anything, when you're running stock settings.
Only if VRAM begins to sit around 110 °C or more, or if you experience:
- Artifacts
- Shut downs
- Driver crashes
- Coil whine which change under load
... Then we will need to thermal pads or airflow
But your values are completely normal.
- VRAM at 100-102 °C = normal and safe
- GPU at 52 °C = excellent
- Hotspot at 74 °C = fully as it should be.
- No undervolt necessary
- The card is built to run like this long term
If you have further questions or need help, then you're welcome to contact us again;
You can do that by responding to this e-mail.
Have a nice day!
At the very least it sounded AI assisted, since the VRAM used in the example for stating thermal tolerance was not the actual type the 9070 xt uses. The 9070 xt uses normal GDDR6. GDDR6X would have been a great point, if it was actually the type being used in the 9070 xt, since it usually is known for running hot.
Edit:
It should be noted that this is a model with SK Hynix GDDR6. A model with samsung memory have been tested by hardware unboxed in their 9070 xt roundup where their model stabilized at 74 °C VRAM temperature. That test is not as intensive in terms of read write operations, but should still be somewhat intensive.