GPU generating a dark display during speedup
GPU generating a dark display during speedup
Hi, I wanted to explore how much my computer could be overclocked using the program Evga Precision X 16. My graphics card is a GTX 960 WindForce 4 GB. I attempted to overclock it on a game for seven days, aiming for maximum graphics performance. It performed quite well at first, around 40-60 frames per second. However, when I reached about 1533 MHz, the temperature climbed to only 60°C. After playing for a short time, my screen went black and the overclocking stats reset to around 1444 MHz. I also tried increasing my memory clock, which came up to about 3545 MHz during the overclocking process. I have questions about what memory clock actually does and whether it needs to be overclocked as well. Additionally, I’m curious about why the screen went black and my settings reset. My ideas were that it reached its maximum power, possibly at 100% of its capacity, or that something might have corrupted the system and stopped it from continuing.
It suggests you likely exceeded the OC too much for the power limits you had in mind. Unless you took a risky action like increasing core voltage by 200mV, the chance of lasting damage is minimal. Lower the overclocks by 10MHz and check if you encounter another crash. Keep repeating until there are no more crashes, then decrease by another 5-10MHz as a precaution for future wear and to provide extra safety.
When examining your Windows Event Log, you might encounter messages about graphics driver crashes or GPU issues. If such entries appear, it indicates the GPU or drivers identified an error and reset the overclock to default settings to minimize further crashes. In such scenarios, higher frequencies like 1533MHz core or 3545MHz GDDR5 could have been excessive. Overclocking RAM is necessary since graphics rendering demands significant memory bandwidth. However, if the core is frequently occupied by VRAM, increasing its speed won't be beneficial.
InvalidError
If you review your Windows Event Log, you might encounter a message indicating a graphics driver crash or a GPU issue. When this happens, the system reset your settings due to an error detected by the GPU or drivers. The overclock was likely triggered because the GPU or drivers identified a problem and reset it to safe default values to minimize future crashes. In such cases, the 1533MHz core or 3545MHz GDDR5 might have been too high.
Overclocking RAM is unnecessary since graphics rendering heavily relies on memory bandwidth. If the core is already busy with VRAM, overclocking offers no benefit.
I checked the Event Viewer and found only a warning: "Display driver nvlddmkm stopped responding and has successfully recovered." (event ID 4101). It doesn<|pad|>, it appears to have stopped without a crash.
Stopped responding is now a distinct form of crash: rather than terminating due to an illegal instruction, it simply halted operation. In either case, an internal driver or hardware issue led to the system reaching a dead state.
InvalidError :
Stopped responding is now a distinct crash type: instead of terminating due to an illegal instruction, it simply froze. Regardless, this indicates an internal driver or hardware issue that led the driver to enter a dead state.
Does this imply something serious? Can I still attempt overclocking at lower frequencies? Is there a solution to fix this, or is it just because I pushed the system too far?
It suggests you likely exceeded the OC too much for the power limits you had in mind. Unless you took a risky action like increasing core voltage by 200mV, the chance of lasting damage is minimal. Lower the overclocks by 10MHz and check if you encounter another crash. Keep repeating until there are no more crashes, then decrease by another 5-10MHz as a precaution for future wear and to provide extra safety.