Skip to main content

Hardware Optimization

This Distro Spin is specifically tuned for the Framework Laptop 13 equipped with the AMD Ryzen™ AI 300 Series (Ryzen™ AI 9 HX 370).

Target Hardware Manifest​

  • System: Framework Laptop 13
  • Processor: AMD Ryzen™ AI 300 Series - Ryzen™ AI 9 HX 370 (Zen 5, RDNA 3.5, XDNA 2 NPU)
  • Memory: 96GB (2 x 48GB) DDR5-5600 (UMA maximized via BIOS for local LLM inference)
  • Storage: 2TB WD_BLACK™ SN850X NVMe™ M.2 2280
  • Display: 2.8K (2880x1920, 3:2 ratio - Wayland 150% fractional scaling defaults)
  • Expansion I/O: USB-C, USB-A, HDMI (3rd Gen), and Ethernet.

Kernel Optimizations​

The following kernel arguments are applied by default to ensure optimal performance and stability on the Framework 13:

  • amd_pstate=active: Enables the modern AMD P-state driver for superior power management and performance scaling.
  • amdgpu.sg_display=0: Addresses specific display flickering and stability issues with the 2.8K high-resolution panel.

Display & Scaling​

The Framework 13's 2.8K display (2880x1920) provides a high pixel density.

  • Default Scaling: We default to 150% fractional scaling on Wayland to provide a balanced workspace.
  • Optimization: We use GSettings overrides to ensure consistent text scaling across the GNOME environment.

Memory Management​

With 96GB of DDR5-5600 RAM, this build is optimized for large-scale local LLM inference.

  • UMA Maximization: It is recommended to maximize the UMA buffer in the BIOS to allow the integrated RDNA 3.5 graphics to leverage the full memory pool.
  • ROCm Integration: The OS is configured to pass this memory pool directly into the Quarantined AI environment.