Scalable Analysis via Machine Learning: Predicting Memory Dependencies Precisely

Abstract

Using Machine Learning to yield Scalable Program Analyses

Program Analysis tackles the problem of predicting the behavior or  
certain properties of the considered program code.  The challenge lies in
determining the dynamic runtime behavior statically at compile time. 
While in rare cases it is possible to determine exact dynamic properties 
already statically, in many cases, e.g., in analyzing memory dependencies, 
one can only find imprecise information. To overcome this, we apply 
Machine Learning (ML) techniques which are particularly suited for this 
task.  They yield highly scalable predictors and are safely applicable when 
erroneous predictions merely have an impact on program optimality but
not on correctness.

In this talk, I present our approach to mitigate the impact of the memory 
gap. Over the last decade, computer performance is often dominated
by memory speed, which did not manage to keep pace with the ever
increasing cpu rates. We consider novel speculative optimization
techniques of memory accesses to reduce their effective latency. 
We trained predictors to learn the memory dependencies of a given pair 
of accesses, and use the result in our optimization do decide about the 
profitability of a given optimization step.