Yinnian Lin‘s paper titled “Accelerating Triangle Enumeration on FPGA-CPU Heterogeneous Platforms”, focusing on improving triangle enumeration on emerging heterogeneous hardware, have been accepted by VLDB Journal 2025.

Triangle enumeration is fundamental in graph analysis, which is the basis for computing various graph metrics like k-truss and clustering coefficients. Moreover, many real-world graph analysis applications such as spam detection, link recommendation, and community detection can be done with triangle enumeration. We present TEAF, a triangle enumeration acceleration system optimized for CPU-FPGA heterogeneous platform featuring adaptive strategy for intersection and a series of hardware-level optimization. Experiments show that TEAF prevails over the previous systems.