An option pricing calculation hardware accelerator, accelerator card and computer equipment

Abstract

The invention discloses an option pricing calculation hardware accelerator, an accelerator card and computer equipment. The option pricing calculation hardware accelerator of the invention includes a Gaussian random number generator, a multiplier m2, an adder a2, an EXP module, a multiplexer MUXA, Multiplex its MUXB, multiplier m5, subtractor s0, comparator, delay module delay, accumulator, shifter and multiplication array, through the combination of the above components, it is possible to simulate a Monte Carlo in M ​​clock cycles The path is iterated to complete the prediction of the option execution price SM once. In order to reduce hardware resource consumption and circuit complexity, 64-bit floating-point operations are converted into fixed-point operations; the accelerator card is a board including the aforementioned hardware accelerator, and the computer equipment is equipped with the aforementioned option pricing calculation hardware accelerator. The present invention can realize Monte Carlo option pricing calculation hardware acceleration without pause, and the entire calculation process is fully streamlined, and has better performance and energy efficiency ratio compared with CPU and GPU implementations under the same process.

Description

technical field [0001] The invention relates to a hardware acceleration technology for Monte Carlo option pricing, in particular to an option pricing calculation hardware accelerator, an acceleration card and computer equipment. Background technique [0002] Monte Carlo option pricing is an existing software algorithm such as figure 1 As shown, the calculation process of Monte Carlo option pricing mainly includes two loops: the inner loop (lines 8 to 11) simulates a stochastic prediction path of option prices; the outer loop (lines 6 to 15) calculates the return and converts Returns are accumulated across all paths, then the sum of returns is averaged and discounted (lines 16-17) to get the predicted option price. For the above software algorithm model of Monte Carlo option pricing, using FPGA to implement hardware accelerator is expected to improve the computational efficiency of Monte Carlo option pricing. However, the use of FPGA to implement hardware accelerators is ex...

AI Technical Summary

Problems solved by technology

However, the current use of FPGAs to implement hardware accelerators is expected to improve Monte Carlo option pricing. The hardware accelerators generated by directly converting software algorithms still have a lot of room for optimization, and there are problems with insufficient energy efficiency and performance.

Images