Skip to content

TooManySticks/Cloud_GPU_Price_Performance_Analyzer

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

10 Commits
 
 
 
 
 
 
 
 

Repository files navigation

Cloud_GPU_Price_Performance_Analyzer

This project analyzes cloud GPU offerings by assigning a price/performance score based on key features like VRAM, RAM, vCPUs, storage, and price.

The goal is to provide an easy way to compare and evaluate different cloud GPU options using a standard scoring system.

The score is calculated using a weighted formula that normalizes each feature to a 0-1 scale and combines them based on configured weights.

Higher scores represent better price/performance. The final score is also converted to a letter grade A-F for easy interpretation.

Features Loads GPU data from an Excel spreadsheet Config driven weights and min/max values for normalization Calculates normalized and weighted score for each GPU Converts score to letter grade Includes tests for validation Usage The main entry point is main.py. Simply run:

Copy code

python main.py This will:

Load data.xlsx Process and normalize features Calculate scores Output score and grade for each GPU The notebook GPU-Analysis.ipynb provides examples of analyzing the processed data.

Resources data.xlsx: contains sample GPU data config.py: weights and min/max values for normalization main.py: main script utils.py: functions for loading, normalizing and scoring GPU-Analysis.ipynb: sample analysis notebook Next Steps Potential enhancements:

Expand to more GPU models Add more features like PCIe vs SXM Build web interface for easy lookup Automate data updates via web scraping

Credits Created by Ben Sutton + AI (ChatGPT & Claude2) as a sample project to demonstrate GPU analysis in Python.

About

Provides an alphabetical (A-F) or numerical value (1-100) for various Cloud GPU offerings based on price and performance features. Only for 1xA100 80GB GPU and eventually scale out to many NVIDIA GPUs (V100s, H100s, GH200s, MI200, etc.).

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages