NVIDIA has introduced a new tool which will make it simpler for customers to overclock its upcoming 20-series GPUs that are going to start shipping next week. The NVIDIA Scanner tool aims to eliminate the trial and error phase from the process of overclocking, bringing one-click overclocking to customers.
The trial and error phase involves adjusting variables to find the sweet spot. If the clock speed is increased too much the system can become unstable and crash. The process usually involves increasing the clock speed and running intensive workloads to make sure that everything is working as it should be. For increased stability, the GPU voltage is then adjusted. It’s a bit more complex than that but you get the idea.
NVIDIA Scanner will make things simpler as customers won’t have to go through this process on their own. It will be automated as Scanner will incrementally increase the clock speed and voltage of the GPU based on the card’s capabilities and ensure that it runs properly when overclocked. It will thus be able to figure out the limits that the card can handle without making the system crash and reboot. NVIDIA says that the entire process can be done in 20 minutes with Scanner.
Scanner only performs overclocking on the GPU so those who are interested in memory overclocking will have to stick with the old methods. It’s also going to be possible to use conventional methods for GPU overclocking even with Scanner.
The tool will work with NVIDIA’s RTX 2080 GPUs which the company has developed with overclocking in mind. NVIDIA does have plans to make the tool available on older generation of GPUs as well.