README.virtual_machine.md

Running scrcpy in a Virtual Machine

When running inside a virtual machine (in the background so you can still use your mouse on the host), it is recommended to limit resources for smoother performance:

• Memory: 8 GB RAM
• CPU: 2 cores

---

1. (Optional) Enable GPU Rendering

Force hardware GPU rendering instead of software:

set SDL_RENDER_DRIVER=direct3d

---

2. Run scrcpy with Low CPU Usage

You don’t need high FPS for automation or vision tasks. Run scrcpy with these options so you save resources:

scrcpy --max-fps 10 -b 2M --video-codec=h264 --no-audio

Explanation of flags:

• --max-fps 5–15 → Limits FPS, fewer frames to decode and draw.
• --no-audio → Disables audio capture/decoding thread.
• --video-codec h264 → H.264 is cheaper to decode than HEVC/AV1 in a VM.
• -b 2M → Reduces bitrate (lower bandwidth and CPU use).

---

3. Networking and USB Setup

• Set Network Adapter to Bridge mode so the VM is on the same network as your host.
• Enable USB forwarding so scrcpy inside the VM can access your phone.

---

4. Using the GPU in host while you run in Virtual Box

The file: server\main_inference.py, is made to expose endpoints to process YOLO and OCR stuff. So if in your VM you set the IP of your running host main_inference.py, and in settings (in VM) you set: USE_EXTERNAL_PROCESSOR = True EXTERNAL_PROCESSOR_URL = "http://YOUR HOST IP:8001"

You will be able to have the best of both worlds: a separated mini computer running and playing the game, and that mini computer will be using your main GPU and host power. This may be a little advanced stuff. Maybe I document more details about this later...

5. My Virtual Box configurations

6. Usage Tips

• Follow the same scrcpy installation instructions.
• You can minimize the VM window while scrcpy runs in the background.
• This lets you continue working on your host machine while scrcpy stays active inside the VM.

---