Graph DBMSs have recently become very popular for various analytical and ML tasks. Traditionally Graph DBMSs very viewed with some skepticism due to their poor performance and scale-out characteristics. For example, the most popular open source Graph Database Neo4j in practice doesn't scale to multiple nodes. But newer systems have arrived, such as TigerGraph and Nebula, that are designed from the group-up as distributed, massively parallel systems. In this talk I will talk about this new development, what are the key advantages of Graph MPP technology over relational MPP, and propose that in fact the future of Big Data Analytics could in fact be in Graph MPP technology.

Слайды доклада

Видео доклада.

References

1. Why graph DB + AI may be the future of data management. https://www.zdnet.com/article/why-graph-db-ai-may-be-the-future-of-data-management/
2. The future of big graphs: A community view on graph processing systems: https://dl.acm.org/doi/abs/10.1145/3434642
3. TigerGraph Scalable Graph Database for the Enterprise. http://www.tigergraph.com
4. Nebula Open Source Distributed Scalable Database. https://nebula-graph.io/