Rate My Professor Richard Membarth

Membarth's research specializes in parallel computer architectures and programming models, emphasizing automatic code generation for architectures from embedded systems to high-performance computing installations, applied to image processing, computer graphics, scientific computing, and deep learning. Notable publications include "GPUs All Grown-Up: Fully Device-Driven SpMV Using GPU Work Graphs" (2025, with Wildgrube et al.), "Exploiting Data Redundancy in Im2Col Convolutions" (2025, with Pfaller), "Exploration of Efficient Computation for Trajectory Planning via Fixed-Point Arithmetic" (2025, with Xu), "AnyQ: An Evaluation Framework for Massively-Parallel Queue Algorithms" (2023, with KenZel et al.), "XEngine: Optimal Tensor Rematerialization for Neural Networks in Heterogeneous Environments" (2022, with Schuler and Slusallek), and "AnyHLS: High-Level Synthesis with Partial Evaluation" (2020, with Özkan et al.). He is a member of the HiPEAC European Network on High Performance Edge and Cloud Computing and has contributed to conference leadership, including General Chair for High-Performance Graphics in 2019 and Papers Chair in 2020. His projects, such as AnyDSL and Hipacc, advance domain-specific languages and compilers for high-performance applications.