Quick Start

Fast GPU deployment for users with account already configured. Deploy in under 3 minutes.

Recommended Configurations

Choose a configuration based on your use case:

Development & Testing

GPU: RTX 4090 (24GB VRAM) Cost: ~$0.52/hour Best for: Prototyping, small models, testing

Production Training

GPU: H100 SXM5 (80GB VRAM) Cost: Variable (check dashboard) Best for: Large language models, production training

Research & Fine-tuning

GPU: A100 (40GB/80GB VRAM) Cost: Variable (check dashboard) Best for: Model fine-tuning, research workloads

Deploy in 3 Steps

Deploy a GPU instance in seconds

1. Select GPU

Go to app.spheron.ai → Deploy

Choose from recommended configurations or browse catalog:

RTX 4090 for development and testing
A100 for production training
H100 for large-scale LLM work

2. Configure

Region: Closest to your location
OS: Ubuntu 22.04 LTS
SSH Key: Select from your uploaded keys
Review pricing in order summary

3. Launch

Click Deploy Instance and wait about 30 seconds. Copy the SSH command from the instance details panel in the dashboard. The username and port vary by provider.

# Spheron AI - username is ubuntu
ssh ubuntu@your-instance-ip

Verify & Test

Check GPU

nvidia-smi  # Should show GPU model, memory, driver

Quick Tests

# Test CUDA
nvcc --version
 
# Test PyTorch (if installed)
python3 -c "import torch; print(torch.cuda.is_available())"
 
# Monitor GPU
nvidia-smi -l 1

Install ML Stack

# Quick install for common libraries
pip install torch torchvision transformers accelerate bitsandbytes
 
# Or use conda
conda install pytorch torchvision torchaudio pytorch-cuda=12.1 -c pytorch -c nvidia

Advanced Options

Startup Scripts

Automate setup with cloud-init scripts. Add during deployment to:

Install dependencies on first boot
Configure environment
Clone repositories
Setup monitoring

See Startup Script Examples for templates.

Managing Costs

Terminate instance when done:

Go to instance dashboard → Click Terminate
Stops all charges immediately
All data permanently deleted

Tip: Use Reserved GPUs for long-term work to save 30-50% on costs.

What's Next?

Deploy AI Models

Deploy LLMs - Run Qwen, Chandra OCR, and more
AI Nodes - Gonka AI, Pluralis

Advanced Setup

Jupyter Notebook - Browser-based development
VS Code Remote - Remote development environment
Startup Scripts - Automate configuration
Templates & Images - Copy-ready startup scripts for common stacks

Platform Features

API Reference - Automate deployments with API
Reserved GPUs - Save 30-50% with reservations
Security Guide - Best practices

Troubleshooting

SSH connection issues:

Verify correct SSH key uploaded: Check User Settings
Try explicit key: ssh -i ~/.ssh/id_ed25519 user@ip
See SSH Guide for detailed help

GPU not showing:

Wait 30s after deployment (drivers loading)
Run nvidia-smi to verify
Reboot if needed: sudo reboot

Deployment failed:

Check account balance has sufficient credits
Try different region (some may be at capacity)
Contact support via Discord

Need Help?

Community: Discord - Fast help from community and team
Docs: All Guides - Complete documentation
Support: General Info - Official channels