CARTS Benchmarks

Feb 1, 2025 · 1 min read
projects

This repository supports the empirical side of the CARTS research line.

It packages benchmark workloads and evaluation harnesses used to study how compiler-visible contracts can expose schedulable structure and improve execution on modern systems.

Authors
Ph.D. Candidate in Electrical and Computer Engineering
I design execution-contract systems that make State, Dependency, and Effect compiler-visible across compilers, runtime systems, and agentic AI workflows. My work spans LLVM and MLIR pipelines, OpenMP and event-driven runtimes, and typed orchestration systems for heterogeneous and distributed execution.