GPU-based maximum k-truss discovery algorithm

The invention discloses a GPU-based maximum k-truss discovery algorithm, and the method comprises the following steps: (1) firstly inputting graphs G and k, solving k-core in the graph G, copying the solved k-core into a video memory, and initializing a set D into a null set; (2) employing a plurality of threads for processing in the video card; (3) in each stream multiprocessor, executing an operation on the vertex v; (4) copying a neighbor of the vertex v to a shared memory corresponding to the current stream multiprocessor; (5) distributing a thread t for each neighbor n in the vertex for processing; (6) when all neighbors of v are processed, summing the values of the registers in the step (5). According to the method, the maximum k-truss discovery algorithm based on the GPU is adopted, the parallelism of the CPU is fully utilized, the operation speed is high, and the operation cost is low. 本发明公开了一种基于GPU的极大k-truss发现算法，步骤如下：(1)首先输入图G、k，求出图G中的k-core，并将求得k-core复制到显存中，初始化一个集合D为空集；(2)在显卡中使用多个线程进行处理；(3)在每个流多处理器中