Skip to content

Choose GPU or CPU for speech recognition

Select whether the whisper.cpp speech-to-text engine uses your system's GPU or CPU for transcription. GPU processing is faster but requires VRAM; CPU processing works on any system but is slower.

Before you start

  • Copy Assist must be open: View > Copy Assist (Ctrl+Shift+T)
  • A radio must be connected

Steps

  1. Open the Copy Assist settings dialog by clicking Settings in the Copy Assist panel.
  2. Locate the Compute device combo box.
  3. Select either GPU or CPU from the dropdown.
  4. Close the settings dialog.

The new compute device is used the next time you enable transcription.

What each control does

Control Default Valid range Behavior
Compute device GPU (CUDA/Metal) GPU / CPU Selects whether whisper runs on GPU (faster, needs VRAM) or CPU (slower, works everywhere)

Tips

  • If you see high backlog values (amber/red) with GPU selected, try switching to a smaller model tier first before falling back to CPU.
  • On systems with limited GPU memory (e.g., integrated graphics), CPU mode may be more stable.
⚡ Athena AetherSDR Assistant
Hi! I'm Athena, the AetherSDR AI assistant. Ask me anything about installation, configuration, or troubleshooting.