TurboLynx is a fast
analytical
graph database
Across a knowledge graph of organizations and their related entities, find the 10 most-connected entities reachable within three hops of any organization.
-- Multi-hop traversal on a schemaless knowledge graph
MATCH (s:Entity)-[:RELATED*1..3]->(t:Entity)
WHERE s.type = 'Organization'
RETURN t.label, count(*) AS n
ORDER BY n DESC LIMIT 10Why TurboLynx?
Schemaless
Nodes and edges carry different attributes with no predefined schema. Handles DBpedia's 2,796 unique attribute types without ETL or migration.
Read moreFast
Extent-based columnar storage, SIMD-vectorized operators, and a graph-aware Cascades optimizer that pushes joins below UNION ALL per graphlet.
Analytical
Group-by, aggregation, and multi-hop traversal in a single Cypher query. Outperforms graph databases on traversal and RDBMSes on graph-shaped joins.
Read moreEmbedded
Like DuckDB and Kuzu — link the library, start querying. No server, no daemon, no IPC. Embed TurboLynx directly via the C API.
Read moreInstall TurboLynx
Start using TurboLynx in your environment.
git clone https://github.com/postech-dblab-iitp/turbograph-v3 && cd turbograph-v3 && cmake -B build && cmake --build build#include "main/capi/turbolynx.h"
int64_t conn = turbolynx_connect("/path/to/workspace");
if (conn < 0) { /* handle error */ }