This paper proposes a GPU-based high-speed signal generation algorithm for effectively simulating the operational environment of the Ex-Core Neutron Flux Monitoring System (ENFMS), which is essential for advanced reactor systems such as Small Modular Reactors (SMRs). Accurate and rapid modeling of neutron, gamma-ray, and electrical noise signals is essential for reliable nuclear fuel monitoring and early anomaly detection. Although conventional CPU-based sequential simulation methods provide precise results, they become impractical under high reactor power conditions or extended simulation durations due to excessive computational demands. To resolve these limitations, we developed a parallel computing framework optimized for high-performance task distribution between CPU and GPU resources. Experimental results demonstrate that the proposed GPU-based implementation reduces elapsed times by up to 99.57 %, 99.43 %, and 98.54 % compared to CPU implementations using MATLAB, Python, and C, respectively. Therefore, the proposed GPU-based parallel algorithm significantly enhances feasibility of realistic and efficient ENFMS simulations, contributing to accelerated development and validation of digital and compact SMR systems.