High-Performance Outlier Detection Algorithm for Finding Blob-Filaments in Plasma

Magnetic fusion could provide an inexhaustible, clean, and safe solution to the global energy needs. The success of magnetically-confined fusion reactors demands steady-state plasma confinement which is challenged by the edge turbulence such as the blob-filaments. Real-time analysis can be used to monitor the progress of fusion experiments and prevent catastrophic events. We present a real-time outlier detection algorithm to efficiently find blobs in fusion experiments and numerical simulations. We have implemented this algorithm with hybrid MPI/OpenMP and demonstrated the accuracy and efficiency with a set of data from the XGC1 fusion simulation code. Our tests show that we can complete blob detection in two or three milliseconds using Edison, a Cray XC30 system at NERSC and achieve linear time speedup. We plan to apply the detection algorithm to experimental measurement data from operating fusion devices. We also plan to develop a blob tracking algorithm based on the proposed method.