RTX 2080 SUPER Mod Doubles VRAM to 16GB, Gains 28% FPS

A hardware modification to the GALAX GeForce RTX 2080 SUPER effectively doubles its video memory capacity, offering a potential solution for games that exceed standard 8GB limits. This upgrade allows users to bypass VRAM bottlenecks that previously caused frame rate drops in memory-intensive titles. The modification is particularly relevant for players who encounter stuttering or low performance in modern games that demand more than 8GB of video memory.

Hardware swap bypasses memory bottlenecks in Resident Evil Requiem

The project involves a GALAX GeForce RTX 2080 SUPER HOF graphics card built around the TU104 GPU architecture. Modders from the Paulo Gomes team replaced all eight original 1GB Samsung GDDR6 memory modules with 2GB chips. This physical swap increases the total memory from 8GB to 16GB while keeping the original GPU core and memory bandwidth specifications unchanged.

• Original VRAM: 8GB
• Modified VRAM: 16GB
• Memory Modules: 2GB chips replacing 1GB modules
• GPU Architecture: TU104
• Test Game: Resident Evil Requiem

The team installed a physical switch on the card to toggle between the original 8GB mode and the new 16GB configuration. They also adjusted the memory strap settings to accommodate the larger 2GB chips. In testing, the 16GB mode allowed memory allocation to rise to approximately 11GB, up from the full 8GB limit in the original configuration.

Performance testing in Resident Evil Requiem with maximum settings showed a significant improvement in the 16GB mode. The original 8GB mode yielded 24 to 26 FPS, while the 16GB mode achieved 32 to 34 FPS, representing a 28 percent gain. The team noted that the mod may not be stable for general use due to display issues at low GPU clocks, which required a modified NVIDIA driver configuration to resolve.

This modification demonstrates that VRAM capacity can be a limiting factor for the RTX 2080 SUPER in specific gaming scenarios. The upgrade provides a tangible performance boost by preventing memory-related stalls, though it requires technical expertise to implement. Users considering this mod should be aware of the potential stability issues and the need for custom driver setups.