I increased the GPU's speed but noticed the voltage drops when under stress, causing crashes.
I increased the GPU's speed but noticed the voltage drops when under stress, causing crashes.
So I increased my GTX 650 Ti performance with MSI afterburner
I raised the core voltage by +100 (to 1.125v),
+200 for the core clock (1124 Mhz),
and +350 for the memory clock (3055 Mhz).
During the Unigine Heaven Stress test, I watched the on-screen readings.
I saw the overclock stayed steady at 1.125v for a short time, then it dropped to 1.112v and remained stable for a while. Later it fell again to 1.10v, which caused the GPU driver to crash after about three seconds.
I checked the "Force constant voltage" option in MSI afterburner settings, but I’m not sure why this happened—it seems like the GPU is intentionally undervolting without any reason.
The temperatures stayed around 73ºC maximum.
I’m using the newest drivers and Windows version.
I really need someone to assist me.
Thanks! Please let me know if you need more details.
The temperatures on your GPU might not exceed 73 degrees, but the VRMs could be overheating, leading to a forced shutdown and crash.
The issue seems to stem from voltage adjustments, but the problem persists even when lowering the overclock. If VRM temperatures were the cause, it shouldn't take much longer to undervolt since this setting is lower than the previous one.
I didn't purchase these items; I just had some leftover components from old motherboards and power supplies. I also had a large thermal pad that could be cut into smaller pieces as needed. I bought this one a long time ago, so you should be able to find similar parts easily. Additionally, I changed the cooler. Here are some pictures: https://imgur.com/a/OeVQd
SyMix_KD shared their experience about acquiring these components. They mentioned they didn’t purchase them directly but sourced some from old motherboards and power supplies. They also noted having a large thermal pad that could be cut into smaller pieces as needed. Additionally, they switched the cooler and provided pictures for reference.
The issue is keeping it in place, I actually had to flip my case upside down since the heatsink is too heavy and wouldn't rest properly on the dye, causing temperatures to spike when I ran anything. How can you find a GPU water block that fits every single GPU? I'm a bit curious.
The issue is keeping it in place, I actually had to flip my case upside down since the heatsink was too heavy and wouldn't rest properly on the dye, causing temperatures to spike when I ran anything. How can you find a GPU water block that fits every single GPU? I'm a bit curious.