An Energy Efficient FPGA Accelerator for Monte Carlo Option Pricing with the Heston Model

Abstract

Today, pricing of derivates (particularly options) in financial institutions is a challenge. Besides the increasing complexity of the products, obtaining fair prices requires more realistic (and therefore complex) models of the underlying asset behavior. Not only due to the increasing costs, energy efficient and accurate pricing of these models becomes more and more important. In this paper we present -to the best of our knowledge -the first FPGA based accelerator for option pricing with the state-of-the-art Heston model. It is based on advanced Monte Carlo simulations. Compared to an 8-core Intel Xeon Server running at 3.07GHz, our hybrid FPGA-CPU-system saves 89% of the energy and provides around twice the speed. The same system reduces the energy consumption per simulation to around 40% of a fully-loaded Nvidia Tesla C2050 GPU. For a three-Virtex-5 chip only accelerator, we expect to achieve the same simulation speed as a Nvidia Tesla C2050 GPU, by consuming less than 3% of the energy at the same time.

A Domain Specific Language for accelerated Multilevel Monte Carlo simulations

“…Multilevel Monte Carlo has shown promising results with FPGA accelerated designs [7]. Combining multilevel optimisations with mixed precision improvements has also shown strong results [2].…”

An Environment for Rapid Derivatives Design and Experimentation

“…De Shryver et al [18] found a Xilinx Virtex-5 FPGA accelerated system to be slower, but 2.5 times more energy efficient, than a Tesla C2050 GPU accelerated system for a one dimensional Heston (stochastic volatility) option pricing model. The authors predict that 3 such FPGAs would give similar speeds to the GPU while consuming only 3% of the energy, if the full calculation were performed on the FPGAs.…”

Energy-efficient FPGA implementation for binomial option pricing using OpenCL

“…Still, the energy cost of acceleration is starting to be taken into account as a key performance criterion, along computing times and accuracy. Following this trend, de Schryver, et al [4] have presented a benchmark to compare option pricing accelerators between each other, and not only between an accelerator and a reference software. They define an option pricing accelerator as:…”