Benchmarks Vs TigerGraph


I’d like to know if you have some benchmarks I can check out comparing Dgraph to TigerGraph, especially regarding latency and volume.


Here’s a feature/spec comparison: Dgraph vs TigerGraph – Dgraph | GraphQL Cloud Platform

I don’t believe there has been any benchmarking performed against tigergraph.

Yeah I’ve seen it, I was more looking for benchmarks but thanks

I don’t think there is, but TigerGraph uses node-based sharding like neo4j et al, (rather than predicate based), so even if you use their affinity-based partitioning (which they recommend against) I expect they will have cases where large data sets in a distributed environment cause excessive network communications when trying to navigate links among many nodes during query processing.

In contrast (and I think uniquely) Dgraph uses predicate-based sharding, which allows a query to execute using a fixed number of network calls between nodes (about equal to query join count). See the “developers in sf who use vim” example here: Database sharding: How to scale a graph database - Dgraph Blog .

1 Like