Which models are you running on 32Gb VRAM (16+16) and 128Gb RAM?
A user recently upgraded their consumer tower, now featuring dual RTX5080s, providing 32Gb VRAM (16+16), and 128Gb DDR5 RAM, powered by an AMD Ryzen 7 7700X CPU. They are seeking recommendations for models that can be run on this specific setup, particularly interested in llama.cpp settings and optimizations from others with similar configurations.
- Published
- Sep 7, 2026, 10:38
- Source type
- Dev community
- Tier
- Community
- Source status
- Healthy
Times shown in UTC
More details
I finally pulled the plug and got a second 5080 and 64GB DDR5 RAM to add to my existing 64. Now I have dual RTX5080s and 128Gb DDR5 RAM on a consumer tower with a AMD Ryzen 7 7700X CPU (so I guess sadly dual channel memory access for now).
Which models are you running on a similar setup with 2x 16Gb VRAM and 128Gb RAM? Extra points if you provide your exact llama.cpp settings and optimizations.