A powerful extension for ComfyUI that enables distributed processing across multiple GPUs and machines. Speed up your image generation and upscaling workflows by leveraging all available GPU resources in your network.

Features

Distributed Processing - Split workloads across multiple GPUs automatically
Distributed Upscaling - Tile-based upscaling with intelligent work distribution
Automatic Worker Management - Launch and monitor workers from the UI
Network Support - Use GPUs across different machines on your network
Real-time Monitoring - Track worker status and performance from the UI
Easy Configuration - JSON-based configuration with UI controls
Memory Management - Built-in VRAM clearing and optimization

Requirements

ComfyUI installation
PyTorch with CUDA support
Python 3.8+

Installation

Clone this repository into your ComfyUI custom nodes directory:

cd ComfyUI/custom_nodes
git clone https://github.com/yourusername/ComfyUI-Distributed.git

Restart ComfyUI - the extension will create a default configuration file on first load

Quick Start

Adding Local Workers

Open the Distributed GPU panel (sidebar on the left)
Click "Add Worker" in the UI
Configure your worker:
- Name: Descriptive name for the worker
- Host: localhost for same machine
- Port: Unique port number (e.g., 8189, 8190...)
- CUDA Device: GPU index from nvidia-smi
- Extra Args: Optional ComfyUI arguments
Save and optionally launch the worker

Adding Remote Workers

Follow the same steps as above, but:
Add --listen to the 'Extra Args' field ⚠️ Required!
Open the worker port in your firewall
Launch ComfyUI on the remote machine
Add the worker using the remote machine's IP address

Tip

: If the remote machine has multiple GPUs, create multiple workers with different CUDA devices and ports

Configuration Tips

Setting	Description	Example
CUDA Devices	Use `nvidia-smi` to see GPU indices	0, 1, 2...
Ports	Each worker needs a unique port	8189, 8190...
Extra Args	Additional ComfyUI arguments	See below

Common Extra Args:

--lowvram - For GPUs with less memory
--listen - Required for remote workers
--highvram - For GPUs with 12GB+ VRAM

Nodes

Distributed Collector

Collects and combines results from distributed processing

Usage: Place after any image generation node to enable distributed processing. Works automatically when workers are enabled.

Distributed Seed

Ensures unique seeds across distributed workers for varied generation

Usage: Connect to any seed input. Each worker automatically receives an offset seed for variety.

Ultimate SD Upscale Distributed

Distributed version of Ultimate SD Upscale that processes tiles across multiple GPUs

Usage:

Upscale your image with a regular upscale model (ESRGAN, etc.)
Feed the upscaled image into this node
Configure tile settings based on your VRAM
Enable workers for faster processing

UI Features

Distributed GPU Panel

The control center for your distributed setup:

Feature	Description
Worker Status	Real-time status indicators
Launch/Stop	Control individual workers
Clear Memory	Free VRAM on all workers
Interrupt	Stop current processing
Auto-detect IP	Find best network interface
Worker Logs	View real-time logs

Workflow Examples

Basic Distributed Generation

Your Workflow → Distributed Seed → KSampler → VAE Decode → Distributed Collector → Output

Create your normal ComfyUI workflow
Add Distributed Seed → connect to sampler's seed
Add Distributed Collector → after VAE Decode
Enable workers in the UI
Run the workflow!

Distributed Upscaling

Load Image → Upscale Model → Ultimate SD Upscale Distributed → Save Image

Load your image
Upscale with ESRGAN or similar
Connect to Ultimate SD Upscale Distributed
Configure tile settings
Enable workers for faster processing

Troubleshooting

Common Issues

Workers won't start

Check ports are not in use: netstat -an | grep 8189
Verify CUDA device exists: nvidia-smi
Check ComfyUI path in worker logs

"Worker not managed by UI" message

Worker was started outside the UI
Stop the worker manually and use the UI to relaunch

Images not combining properly

Ensure all workers have the same models loaded
Check that custom nodes are installed on all workers

Network connection issues

Check firewall settings for required ports
Verify master IP is accessible: ping 192.168.1.100
Ensure same ComfyUI version on all machines
Ensure ComfyUI-Distributed is installed on remote workers

Debug Mode

Enable detailed logging for troubleshooting:

UI Method: Settings → Debug Mode → Enable
Check console output for detailed logs

Development

This project is under active development. Contributions are welcome!

Planned Features

Remote worker control via SSH
View remote worker logs in UI
Improve worker timeout logic
Authentication for network security
Automatic model synchronization

Made with ❤️ for the ComfyUI community

Report Issues • Request Features