15-745 Spring 2011
Class Projects

Here are the class projects for 2011 (in no particular order):

Applying Machine Learning to Compiler Optimization:

• Predicting Unroll Factors Using Supervised Classification. Nan Li and Julian Shun.
• A classifier for ordering and selection of dependent LLVM code passes. Brendan Meeder, Jamie Morgenstern, and Richard Peng.
• Learning Function Inlining. Yair Movshovitz-Attias and Dana Movshovitz-Attias.

Optimizations to Target Novel Architectures:

• An Evaluation of a Block-Structured ISA. Chris Fallin and Gennady Pekhimenko.
• High Level Synthesis for CoRAM. Gabriel Weisz.

Optimizing Memory Accesses:

• Software Prefetching in LLVM. Nathan Snyder and Q Hong.
• Modeling memory usage patterns of GPU programs written for the NVIDIA GPU architecture. Cyrus Omar and Salil Joshi.
• Interprocedural Register Allocation for LLVM. Benjamin Blum and Nicholas Tan.

Singular Topics:

• Implementation of Decoupled Software Pipelining (DSWP) in the LLVM Compiler Infrastructure. Fuyao Zhao and Mark Hahnenberg.
• Symbolic Execution in Difficult Environments. David Renshaw and Soonho Kong.
• Partial Redundancy Elimination in CMORT. Luke Zarko.

Back to CS745 home page.