Igor Wallossek of Igor's Lab concluded that many GeForce RTX 40-series cards from various board partners are, in his words, a "ticking time bomb" for overheating because of low-quality thermal paste that dries out and pumps out over time.
What Is Confirmed
- Cards across multiple partners have been reported hitting 100°C and above after a period of use.
- The mechanism is thermal interface degradation, not a design flaw in the cooler itself.
- Temperatures typically climb gradually over months, which is why owners rarely notice until performance drops.
What Is Still Unconfirmed
- How many cards are affected, and which specific partner models are worst.
- Whether this materially shortens card lifespan or simply causes throttling.
How to Check Your Own Card
- Install HWiNFO64 or GPU-Z.
- Run a demanding game or a stress test for 20 minutes.
- Watch GPU Hot Spot temperature, not just the core temperature. Hot Spot is the number that reveals paste problems.
- A Hot Spot to core delta above roughly 20°C suggests degraded thermal interface material. Under 15°C is healthy.
Before You Repaste — Try the Free Things
- Clean the dust out. A clogged heatsink causes the same symptom and costs nothing to fix.
- Improve case airflow. A card starved of intake air will run hot no matter how good the paste is.
- Check that the bottom intake fan is not blocked by a sagging card.
MediaVision Technical Analysis
A repaste with quality compound — and fresh thermal pads on the memory and VRM — commonly drops Hot Spot temperatures by 15–25°C on an affected card. It is genuinely effective work.
The honest caveat: opening the card usually voids the warranty. If yours is still in warranty and running hot, claim it rather than opening it. If it is out of warranty, a professional repaste is a far cheaper path than replacement. We do this regularly and we test temperatures before and after so you can see the difference.
