Truth behind Neo4j's "Trillion" Relationship Graph - DZone Big Data

I am on the ISO GQL standard committee and actively contribute to the GQL graph query language standard. I am writing this blog with my two decades of research (my database Ph.D. program training at the University of Florida) and industry development experience (Microsoft, Oracle, Turn Inc., and TigerGraph) in the database area. I have tried my best to make this technical blog consumable by general readers who are likely to be new to the world of databases. In a nutshell, Neo4j took the LDBC benchmark name and the graph/table schema in LDBC-SNB benchmark, generated its own simplified dummy dataset which is useless in real life (no real correlation between entities, no realistic edge or relationship degrees), cherry-picked 4 simple queries out of the 14 IC queries of LDBC-SNB (page 32) to claim and create the illusion that Neo4j can scale and answer global queries, neither of which is true. A distributed and scalable database doesn't need users to know or care about how many machines or shards the system needs.