Scaling energy per operation via an asynchronous pipeline
Download as DOC, PDF0 likes73 views
This document discusses statistical analysis showing a sharp reduction in the rate of energy efficiency improvements in processors over the last 30 years, forming an "asymptotic wall". The authors developed an energy model for asynchronous systems that accounts for scaling realities. Studies of an energy efficient asynchronous pipeline fabricated in 0.6 μm at sub-threshold showed 17 Giga Operations per Joule when fully pipelined, and simulations in a more modern 65 nm process projected a further order of magnitude improvement.
Scaling energy per operation via an asynchronous pipeline
- 1. SCALING ENERGY PER OPERATION VIA AN ASYNCHRONOUS PIPELINE ABSTRACT: Statistical analysis of computations per unit energy in processors over the last 30 years is given that illustrates a sharp reduction in the rate of energy efficiency improvements over the last several years resulting in the formation of an asymptotic “wall” with our dataset; we use the measure of giga multiply accumulates per Joule. We have developed an energy model which takes into account the realities of scaling, specifically for asynchronous systems. Studies of an energy efficient asynchronous pipeline show fabricated results of 17 Giga Operations per Joule in 0.6 μm at sub-threshold when fully pipelined, and simulations at a more modern 65 nm process show a further order of magnitude improvement on that.