Willow Marie Ahrens
I am a postdoc at MIT advised by Saman Amarasinghe, and an incoming professor at Georgia Tech! I am inspired to make programming high-performance computers more productive, efficient, and accessible. My research primarily focuses on using compilers to adapt programs to the structure of data, bridging the gap between program flexibility and data structure flexibility. I’m the author of the Finch array programming language, which supports a wide variety of programming constructs on sparse, run-length-encoded, banded, or otherwise structured arrays.
Sessions
Finch is a Julia-to-Julia compiler which adapts array programs to the sparsity and structure of data automatically. Finch understands array structure through a language of basic loop building blocks called Looplets. This enables new loop optimizations across multiple domains, unifying techniques such as sparse tensors, geometric programming, databases, and lossless compression.
Sparse matrices and tensors are ubiquitous throughout multiple subfields of computing. The widespread usage of sparse data has inspired a multitude of in-memory and on-disk storage formats, but the only widely adopted storage specifications are the Matrix Market and FROSTT file formats, which are both ASCII text-based.
Panel: What's next for SparseArrays.jl