GPU Research Notebook
A mini notebook for researching GPUs: specs, comparisons, pricing and notes.
About This Project
GPU-Research-Notebook is a structured documentation system designed to track GPU specifications, pricing, and availability. It utilizes a standardized file structure to organize research notes, data snapshots, and automated reports, ensuring that all findings are dated and traceable. By maintaining individual folders for every research run, the project prevents data overwriting and preserves historical market snapshots.
The repository leverages Typst for automated PDF report generation, employing shared templates to maintain consistent formatting for data analysis. It includes custom Python scripts for merging batch CSV data, applying proprietary evaluation rubrics, and ranking hardware based on metrics like VRAM capacity, cost, and scarcity. This tool is intended for developers and hardware enthusiasts who require a reproducible, data-driven approach to evaluating GPU suitability for AI workloads.