ABSTRACT
Large graph computing system is a key tool in big data computing. It can be applied to a variety of big data applications, such as social networks, web page search, and protein interactions. However, the unstructured graph data make data access nonuniform, which poses a great challenge for building an efficient large graph computing system. Fortunately, a lot of large graph computing frameworks have been proposed recently to alleviate the above problems. In general, these frameworks can be categorized into single-node in-memory system, distributed shared memory system, and single-node out-of-core system. Besides, there are some other solutions that utilize flash SSD and GPU to speed up large graph computing. In this chapter, we will review these typical large graph computing systems from a system perspective.
