Last month we showcased GreenBoost as an open-source means of augmenting NVIDIA GPU vRAM with system RAM and NVMe storage. This memory tiering solution for NVIDIA GPUs was developed by an open-source developer with a focus on CUDA and allowing larger LLMs to be handled on graphics cards with smaller vRAM capacities. There was a setback to the project due to NVIDIA legal but now the project is going in new form and also has introduced GreenBoost-Proton for helping Linux gaming on NVIDIA hardware.
After the spotlight was turned to GreenBoost, unfortunately, some issues were encountered. The original NVIDIA Forums discussion over it was removed and the original GitLab repository had to be shutdown due to having “nvidia_” in the repository URL. GreenBoost is not developed by or sponsored by NVIDIA.
That “nvidia_greenboost” repository now shutdown and being removed as to not infringe on the NVIDIA trademark.
The developer is now developing the code via this “greenboost” repository on GitLab while also making it more clear that GreenBoost is not affiliated/developed/sponsored by NVIDIA.
With the latest development code over the past month, there is now “Greenboost-Proton” and a Vulkan layer to help in enhancing Linux gaming support on NVIDIA hardware. When making use of this code, it will now report greater video RAM capacities based on the GPU vRAM and system memory capacity to Vulkan. Games that otherwise were hitting issues due to the limited vRAM capacity should now cope better with the memory tiering of GreenBoost.
Those wishing to check out GreenBoost for a better CUDA and gaming experience with “small” graphics cards can find it at its new home.
