Summary
Michael Lehn is a software engineer and applied mathematician with 14 years of experience focused on high-performance numerical computing. He is the developer of FLENS, a C++ scientific computing library, and authored a widely used 13-step GEMM optimization tutorial and teaching framework (ulmBLAS) that guides students from naive implementations to micro-kernels rivaling Intel MKL. His work blends deep mathematical training (Dr. rer. nat. from Universität Ulm and an MSc from USC) with practical low-level performance engineering on x86-64 SSE. Michael teaches and documents what often remains opaque in numerical libraries—how small, targeted changes in a micro-kernel unlock dramatic, cache-aware performance gains. Based in Ulm, Germany, he describes himself succinctly as mathematician, coder, nerd, reflecting a culture of rigorous, hands-on craftsmanship in open-source numerical software.
14 years of coding experience
Master of Science (MSc), Applied Mathematics, Master of Science (MSc), Applied Mathematics at University of Southern California
Dr. rer. nat., Dr. rer. nat. at Universität Ulm
English