A Python-based calculator that models the 3-year Total Cost of Ownership for running LLM inference and fine-tuning on-premises (VMware/NVIDIA) vs. cloud (AWS SageMaker, Azure ML). Designed for enterprise architects making build-vs-buy decisions.
- GPU type and count
- Utilization percentage
- Electricity, cooling, and data center costs
- Cloud instance pricing (on-demand, reserved, spot)
- Breakeven analysis chart
- Cumulative cost graph
- Recommendation summary
🚧 Initial script and Jupyter notebook in development.