Unified Memory

  • Oct. 11, 2025 / Hardware Insights

    What Makes Apple Silicon and Strix Halo Good at Running Local LLMs?.

    For years, the formula for running large language models locally has been simple: get as much VRAM as you can afford. This usually meant building complex, power-hungry desktop rigs with multiple GPUs or hunting for deals on used server hardware. But a new class of hardware, powered by Apple Silicon and AMD’s “Strix Halo” APUs,...

  • Oct. 8, 2025 / How to Run LLMs Localy

    I optimized my Strix Halo for local LLMs: Here are the benchmarks and findings.

    If you’ve gotten your hands on an AMD Ryzen AI Max+ 395 (Strix Halo) system, you already know the raw hardware is impressive. That massive pool of unified LPDDR5x memory is a game-changer for running large models locally. But unlocking its full potential isn’t just plug-and-play. The key to getting the best possible performance lies...