The traditional orbit propagation algorithm needs great power of computation, generally implemented on workstations in Hypercube way, which demands high energy consuming and cost overhead. With the fast development of aerospace science, the requirement for large-scale concurrent orbit propagation computation is emerging and urgent. To issue this problem, an National high-tech plan(863) has supported a project for implementing the numerical orbit propagation parallel and scalable algorithm based on the novel high-performance heterogeneous computing (HPC) systems

To solve the memory wall and power wall inherent in the chip, next-generation high-performance systems are integrated into new architecture features like graphic processor units(GPUs), FGPAs, and other general purpose computing co-processors. By exploiting the  unprecedented computing power in those emerging the heterogeneous/hybrid system architectures,  it aims to yield high-performance and scalability by designing a parallel and trustworthy propagation algorithm on massively parallel processors, with assurance of a high precision compared to traditional algorithm, but producing a high speedup in algorithm performance.