Working to understand the economic and technical foundations of progress in computing
Our work:
The Computation Limits of Deep Learning
This project aims to catalog the deep learning’s voracious appetite for computing power, showing that progress across a wide variety of applications is strongly reliant on increases in computing power.
Compute Trends Across Three Eras of Machine Learning
This project aims to highlight the fast-growing compute requirements for training advanced ML systems of three eras: the Pre Deep Learning Era, the Deep Learning Era and the Large-Scale Era.
The Importance of (Exponentially More) Computing Power
This project aims to show how more powerful computers are improving outcomes across society. For that, we looked across five key application areas: Computer Chess, Computer Go, Weather Prediction, Protein Folding, and Oil Exploration.