Speaker
Description
In high-energy physics experiments, online data processing plays an important role. Positioned between the readout of the front-end electronics and the disk, it reduces the vast raw data to a storable size through fast reconstruction and event filtering. Next-generation experiments such as the Circular Electron–Positron Collider (CEPC) impose even more stringent requirements on online data processing. Heterogeneous computing—coordinating different processors like CPUs and GPUs—can boost online data-processing capability and meet these higher demands. This report presents and describes a heterogeneous online data-processing framework designed for CEPC, aiming to improve online computing power through heterogeneous computing resources and thereby provide solid online data-processing support for the CEPC.
