Short Bio

Tim Lühnen's research interests are centered on GPU performance optimization and architectural modeling. It focuses on improving the efficiency and scalability of GPU kernels through advanced analysis, optimization strategies, and performance-driven design. It also involves developing detailed simulation models to support GPU architecture studies, providing deeper insights into hardware behavior and informing future architectural advancements.

Publications & Conference Talks

  • ClusterSim: Modeling Thread Block Clusters in Hopper GPUs (IISWC’25)
  • Benchmarking Thread Block Clusters (HPEC’24) - Best Paper Award
  • Auto-TBC: Auto-Optimizing CUDA Applications Using Thread Block Clusters (ISC’24)

Contact