Daniel Leeds

About Daniel Leeds

I am primarily interested in systems-level research for efficient deep learning and scientific computing with high performance computing systems (HPC) and wireless-networked systems. Specific research topics include: reduced and mixed precision deep learning; distributed optimization techniques for high-dimensional problems and problems characterized by sparse matrix operations; and workload distribution techniques for minimizing performance variability on large-scale compute environments. I have a strong interest in the mathematics for non-convex optimization problems in deep learning (classical and quantum), efficient attention mechanisms, efficient sparse matrix algebra, and graph network problems (including deep learning with graph structures). I took 2.5 years to complete my B.S. in Data Science from the University of Georgia. I am now a graduate student at Rice University. My programming skills include Python, C/C++, and GPU programming (CUDA). I have experience with cloud platforms including Amazon AWS, Microsoft Azure, and Databricks. All opinions are my own.

Experience

1. Engineering Co-op (Advanced Technology)

   AMD

   Jan 2025 — Present

Education

• The University of Georgia

  Bachelor of Science - BS

• Rice University

  Doctor of Philosophy - PhD

• Rice University

  Master of Data Science - MDS, Machine Learning Specialization