Knock is a three-dimensional phenomenon in which local fuel/air ratio, temperature, and pressure are all contributors. CFD simulation is designed to predict such spatial and temporal conditions and should be able to predict knocking tendency. However, traditional internal-combustion engine CFD simulation approaches use severely reduced fuel chemistry models that are unable to accurately emulate real auto-ignition behavior. Lacking a trustworthy approach to simulating knock, engine designers have been forced to rely on experimental testing, which is both time-consuming and expensive. Recent improvements made in the accuracy of fuel models now enable predictive simulations of the impact of operating conditions and fuel type on auto-ignition. These fuel models require advanced solution techniques to handle the chemistry detail robustly when coupled to CFD. Reaction Design’s CFD package FORTÉ makes use of CHEMKIN-PRO solver technology that can handle the detail. In FORTÉ, detailed kinetics with hundreds of species and thousands of reactions can be used while maintaining practical time to solution. To predict knocking, auto-ignition events in the simulation are monitored by virtual sensors defined in the FORTÉ CFD model and then analyzed with signal-processing techniques that mimic experimental setups during solution visualization. The results are used to calculate knock intensity.