This paper documents a world record achievement in GPU overclocking through extreme hardware modifications applied to an NVIDIA GeForce RTX 2080 Ti graphics card. By implementing systematic power delivery enhancements, extensive VRM modifications, and innovative wire soldering techniques, we achieved a final benchmark score of 11,155 points in the Unigine Superposition 8K benchmark, securing the number-one global position on the HWBOT leaderboard. The methodology involved progressive augmentation of power delivery capacity via direct wire connections to the PCB, bypassing standard power limitations, and optimizing PCIe interface bandwidth. Our results demonstrate that, with proper power delivery infrastructure, the RTX 2080 Ti can reach GPU core frequencies exceeding 3,100 MHz and memory frequencies of 17,000 MHz (effective)—representing a substantial performance increase over stock specifications. This work contributes to the understanding of power delivery constraints in modern GPU architectures and provides empirical evidence for the relationship between power capacity and overclocking headroom.
Zen Revista (Wed,) studied this question.