Supercomputing 26 takes place in Chicago, Illinois, on November 15-20, 2026. This time our popular tutorial “Core-Level Performance Engineering” was accepted as a full-day event on November 15. Jan Laukemann and I have teamed up with Filippo Spiga from NVIDIA to show you how modern CPUs execute  code. We’ll explain pipelining, hardware threading, out-of-order execution, superscalarity, SIMD, and how they are implemented on x86 and Arm-based CPUs (including the brand-new NVIDIA “Vera”). Hands-on exercises use Matt Godbolt’s Compiler Explorer and OSACA, our Open-Source Architecture Code Analyzer, which is integrated with it. If you need to model the in-core performance of code for optimization or co-design, this tutorial is for you.