Do you have issues with your Zotac 3060 Ti crashing?
Do you have issues with your Zotac 3060 Ti crashing?
My Zotac 3060 Ti, approximately three years old, has been experiencing crashes when the GPU temperatures exceed 80 degrees Celsius. I noticed this while playing a UE5 game, which caused the GPU temperature to rise sharply during play. Once the temperature reached about 82°C, the system would crash with an unrecoverable error.
I was able to recreate the issue by stress testing the PC using FURMARK for several minutes—this time the display would freeze when the GPU hit around 83°C, forcing me to perform a hard reset to restore functionality. Running FURMARK twice produced the same outcome.
During routine tasks, there are no problems, but issues start appearing when the GPU temperature reaches 80°C or higher. Is my GPU nearing failure? Should I consider sending it in for replacement? Could another component be causing this problem instead of the GPU?
I attempted to switch to an older driver, but that didn’t resolve the issue. Please let me know what steps should be taken next.
Should i submit it for RMA?
You still have warranty coverage on the card. You might want to check the card for any dirt or debris that could be blocking the cooling assembly. If the card is outside its warranty period, consider replacing the thermal paste and pads on the card to see if that resolves your issue.
System specs:
CPU: 12400F
PSU: Corsair RMe 750
RAM : 32 GB DDR5
MOBO: Gigabyte B760
When sharing a troubleshooting thread, it's standard to provide full system specifications. List them clearly 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 BIOS version for your motherboard at this time.
PSU: Corsair RMe 750
Not a particularly good unit; it's not ideal.
My Zotac 3060, which is about three years old, raises some concerns because the RTX 3000 series is known to experience high transient load spikes. In this case, it might be your GPU that isn't receiving power from the PSU or your case isn't providing enough cooler airflow.
I attempted to switch to an older driver, but that didn’t fix the problem.
Please explain the steps you took and the driver versions you've used before creating this thread.
CPU: 12400F
CPU cooler: CM Hyper 212
Motherboard: B760m Ds3H (with bios version F21)
Ram: 16gb * 2 DDR5 Corsair 5200 MT/s
SSD/HDD: 3* SATA HDD and 1 SATA SSD (512gb)
GPU: Zotac 3060 ti Twin Edge
PSU: Corsair RMe 750 ( received it as a replacement for my Corsair TXM 750 in June 2023, when the latter malfunctioned)
Chassis: Corsair Carbide 400R with 3 intake fans,1 exhaust fan at the rear
OS: Win 11 pro
Monitor: LG 24 inch IPS
When i encountered the WHEA uncorrectable error for the first time(which happens to be today),i was running my gpu with the latest driver ie 581.15-i had installed it earlier today.
I tested the gpu by running furmark(while the new driver was installed)-and it caused the monitor to black out when the gpu temp exceeded 80 C (although it didn't cause the WHEA error this time around).
i retested it by completely removing the new driver using DDU and reinstalling the older driver (v 580.97) that i had been using earlier (when there were no such issues)-but alas,the results were still the same: ie the moment the temp went north of 80 C(in Furmark) ,the screen goes blank and it requires a hard reset to make it work again.
I'm not sure if its indeed a problem with the GPU itself or some other component such as the PSU(which is around 2 years old).The gpu is still under warranty as it was registered for extended warranty with Zotac in 2022. Will i try testing the system with an alternate psu before sending the gpu for RMA?
If you don't have another PSU, it's pointless to spend more. You can return the GPU, and it won't cost anything.
It could be free if the GPU has an issue, as Zotac will handle returns. Typically, you'd send it back to the store, which would then forward it to Zotac.
But it seems like you're dealing with insufficient airflow. *Warning: the cm hyper 212 I've disliked for ten years might be just adequate, but is it far from 'overkill' for the CPU? And Zotac offers a more compact heatsink too.*
Right now, a powerful cooler such as the Thermalright Burst Assassin isn't too expensive. This would immediately lower system temperatures.
If the issue appears only when the card reaches 83°C after nearly three years of near-maximum use, it's likely struggling to maintain stability and starts throttling. When did you begin tracking temperatures until a problem became clear?
Regarding airflow, consider shifting one front intake to the top exhaust to expel more warm air. Warm air naturally rises, making this a straightforward and effective cooling boost. It also draws air upward from the floor.
The problem begins at 83°C. Therefore, ensure your cooling system never reaches that temperature. The Corsair Rme 750 is a solid PSU with a ten-year warranty.
Still, it doesn’t guarantee perfection, but it’s unlikely to be the main cause here.
The RmX model handles transients quietly better, though the difference isn’t crucial.
However, several factors—such as the 'last-gen' CPU cooler, the compact Zotac cooler, and a budget PSU—could collectively result in subpar cooling performance.
I rarely adjusted fan curves since I’ve never experienced good results with a reliable CPU cooler. Starting with default settings is usually fine.
If you still face issues at lower temperatures, consider RMA after trying other fixes.
But I’d be skeptical about the overall 'suboptimal cooling' approach—especially given the Zotac card’s tendency to run hot. It puts you in a tricky situation and leaves uncertainty about their RMA process.
It's actually an updated version of the Hyper 212, specifically called Hyper 212 Spectrum V3.
I understand that the 3060 Ti tends to overheat, but this problem has appeared recently.
I recall testing the card a few months ago with Furmark, but it didn't cause any issues then.
Recently, temperatures have been rising quickly during GPU-heavy games. When the crash happened, the hotspot reached 96°C and the GPU temperature was about 83°C. I usually check my CPU temperatures while gaming using NVIDIA's overlay.
The crash happened while playing a Unreal Engine 5 game (Cronos the new dawn), which is known to put a lot of strain on the GPU.
It might just be a matter of luck, but the warranty is likely to cover it. The thermal solutions can wear out over time, similar to the internal pads and paste. Fans may also slow down or stop working. It’s probably best not to open the card yourself—this should be handled under warranty.