If you need a GraphQL-native database with high performance and fault tolerance for your distributed app, Dgraph is a great option. It's a distributed graph database that's designed to run natively as GraphQL, so you can build scalable apps as fast as possible. Dgraph is fast at queries, can scale to any size automatically with sharding and rebalancing, and is fault-tolerant. It can run on-premises or on AWS or GCP.
Another contender is Grafbase, a unified graph platform that lets developers assemble multiple data sources into a single graph. It comes with modern tooling for building and deploying GraphQL APIs, federation, and advanced security. Grafbase also supports many tools and services and has a deploy-anywhere flexibility, so it's a good option for distributed apps.
If you need a more real-time data platform, check out SingleStore. It can handle transactional and analytical data in one database engine with high-throughput streaming data ingestion for millisecond query performance. SingleStore supports multiple data models and has read replicas for independent compute scaling, which can be very useful for distributed apps that need to process data as fast as possible.