Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Explore differentiable programming for scientific computing in this 45-minute conference talk from FerriteCon 2025, where Valentin Churavy from the University of Augsburg demonstrates how the Enzyme automatic differentiation framework revolutionizes derivative computation in Julia. Learn how Enzyme generates derivatives of complex scientific code patterns—including those with mutation and control flow—directly from LLVM IR, enabling powerful automatic differentiation capabilities. Discover why differentiable programming has become essential in modern scientific computing for optimization, sensitivity analysis, and matrix-free methods. See practical demonstrations of Enzyme's integration with Julia's scientific ecosystem, allowing you to compute derivatives of high-performance simulations without code rewrites. Gain insights into advanced Julia workflows and understand how differentiable programming transforms real-world computational problems in scientific domains.
