Title: Multi-Level Monte Carlo Simulations for Financial Applications
Speaker: Christian de Schryver, University of Kaiserslautern
Abstract:
The financial and insurance business needs to run a lot of simulations for pricing and risk assessment tasks throughout the day, since for many products no closed-form pricing formulas exist. Although for special types other algorithms may converge much faster, Monte Carlo methods are very robust and applicable to a wide range of problems. For the showcase “European option pricing with the Heston model”, this talk will present hardware efficient solutions for random number generation and path simulations on FPGAs, together with detailed speedup and energy numbers. Furthermore I will highlight how an application-level benchmark suite allows us to compare accelerator solutions in a fair way.