跳转至

Cross-Architecture Compatibility Matrix

Participants bring a wide variety of hardware architectures to workshops. Workshop Harness audits technology stack risks prior to the session and provides mandatory fallback paths.


Hardware Chipset & OS Matrix

Architecture / OS Recommended Tool Known Risks Mandatory Fallback Path
macOS Intel Mac (x86_64) Ollama CLI (ollama serve) LM Studio GPU acceleration unavailable / crashes frequently Must provide Ollama CLI fallback guide (docs/18-intel-mac-prep.md)
macOS Apple Silicon (arm64) LM Studio / Ollama / MLX Full Metal GPU hardware acceleration supported MLX (mlx-lm) optional
Windows x86_64 (Intel/AMD) LM Studio / Ollama PowerShell execution policy restriction, WSL2 required for Docker Provide PowerShell bypass script
Windows ARM64 (Snapdragon) Ollama CLI (Native build) Performance degradation under x64 emulation Use Ollama native build
Linux / ChromeOS Ollama CLI No GUI support / Sandbox container Use ollama serve terminal mode & small models

Automated Architecture Detection (check_architecture_compat)

Workshop Harness embeds architecture detection scripts inside every generated workshop project:

  • macOS / Linux: ./scripts/check_architecture_compat.sh
  • Windows: .\scripts\check_architecture_compat.ps1