Saday Sadayappan

P. (Saday) Sadayappan is a Professor at the University of Utah's School of Computing, specializing in high-performance computing and compiler optimizations. His research focuses on performance optimization for parallel systems, particularly for tensor computations, sparse matrix operations, and machine learning workloads. He leads multiple NSF and DARPA-funded projects focused on GPU optimization, tensor computations, and scalable machine learning frameworks.

Research Interests: Dr. Sadayappan's work spans compiler optimizations for high-performance systems, optimization of sparse/dense matrix/tensor computations, scalable machine learning, and algorithm-architecture co-design. His recent projects include developing performance-portable frameworks for tensor applications and optimizing data locality for scientific computing.

Publication Trends: His recent publications (2020-2022) predominantly focus on GPU acceleration of machine learning workloads (especially CNNs), automated I/O complexity analysis, and optimization techniques for sparse matrix/tensor operations. Earlier work (2018-2019) established foundations in GPU code generation for tensor contractions and cache optimization.

Awards and Honors:

• ACM SIGPLAN Most Influential PLDI Paper Award (2018) for A Practical Automatic Polyhedral Parallelizer and Locality Optimizer

Active Grants and Projects:

• NSF: A Comprehensive Framework for Efficient, Scalable, and Performance-Portable Tensor Applications (2022-2027)
• NIH SBIR: Enabling next generation machine learning for large scale image analysis (2023-2025)
• NSF: AI Institute for Intelligent CyberInfrastructure (ICICLE) (2021-2026)
• NSF: Data Locality Optimization for Sparse Matrix/Tensor Computations (2020-2024)
• DARPA SBIR: Performance Portable Framework for Developing Graph Applications (2017-2022)

Teaching: He currently teaches CS 4230/6230 (Parallel and High-Performance Computing) at the University of Utah.