MoonglowRun local Jupyter notebooks on cloud GPUs

Moonglow: GPU-Powered Jupyter Notebooks Without the Hassle

Introduction

Moonglow enables seamless execution of local Jupyter notebooks on remote GPUs, eliminating the need for SSH configuration. Switch from CPU to GPU processing with ease, directly from your preferred IDE.

Key Features

• Instant GPU access: Transition from CPU to GPU in seconds
• Simplified setup: No SSH keys or package management required
• Diverse GPU options: A40s, A100s, H100s, and more available
• In-IDE GPU management: Start, stop, and restart GPUs within your workspace

Use Cases

• Accelerate machine learning model training
• Enhance deep learning research capabilities
• Boost data processing and analysis tasks

Pricing

Personal Plan

• Free to start
• Connect to your Runpod
• Up to 3 notebooks simultaneously
• File storage integration

Enterprise Plan

• Connect to any infrastructure
• Unlimited notebooks
• Team collaboration features
• Custom demo available

Teams

Founded by Leila (ex-Jane Street high-performance infrastructure) and Trevor (former ML researcher at Stanford's Hazy Research Lab), Moonglow aims to simplify GPU-powered machine learning for technical professionals.