Benchmarking Results are Up to One Thousand Times Faster Than Neo4j

REDWOOD CITY, Calif.: #graphanalytics--Rocketgraph, the high-performance, no-code graph analytics platform company, is excited to announce Rocketgraph Performance Labs (RPL), some of its initial benchmarking results, as well as new efforts under way. RPL is dedicated to conducting in-depth benchmarking and performance analytics. With a mission to enhance the speed, usability, and functionality of graph technologies, RPL aims to provide organizations with critical insights that drive productivity and accelerate time to insight.

“At Rocketgraph, our commitment to innovation is unwavering,” said Brock Alston, Rocketgraph’s CEO. “Over the years, we have focused on delivering solutions that not only meet but exceed the expectations of our customers. Our prior work has established Rocketgraph as a frontrunner in graph technology performance, and the establishment of the RPL is a significant step in our journey to redefine industry standards.”

The initial benchmarking results from RPL demonstrate high performance from Rocketgraph’s approach compared to Neo4j.

Ingest Performance: Utilizing Rocketgraph, we achieved ingestion speeds of up to 100 million attributes per second. In contrast, Neo4j's ingestion rates were found to be significantly lower, with capabilities as low as a few hundred nodes or edges per second. This stark comparison highlights Rocketgraph's ingestion performance as being up to 1000 times faster.

BFS Algorithm Performance: The breadth-first search (BFS) algorithm serves as a critical measure of graph system scalability. In independent testing conducted at RPL, Rocketgraph outperformed Neo4j significantly across various vCPU environments, completing queries that took Neo4j days to process in just 40 minutes. Notably, attempts to benchmark Neo4j at 256 vCPUs resulted in machine failure, underscoring the superior scalability of Rocketgraph.

Temporal Triangles Performance: The Temporal Triangles (TT) query serves as an ideal benchmark for property graph databases by identifying directed 3-cycles where timestamps increase monotonically and fall within a specified time window. In our tests at the lab, Rocketgraph was found to be an average of 300 times faster than Neo4j, with the speedup scaling linearly with the number of vCPUs utilized. This significant performance advantage underscores Rocketgraph's ability to handle complex temporal queries efficiently.

What is new at the lab.

Rocketgraph's Performance Labs (RPL) is actively innovating in the field of graph analytics through partnerships with educational institutions and organizations. At Columbia University, we have enhanced a data analysis course, enabling students to start their work in minutes instead of days. Additionally, our collaboration with the Technology Advancement Center (TAC) provides cyber threat analysts with practical training using Rocketgraph. We are also exploring edge computing capabilities with Sourcecode to determine how we can effectively deliver petabyte-scale graphs in these environments.

“In minutes we can ingest millions of flow records, overlay node-type and register-address metadata, and pivot from high-level topology to coil-level anomalies,” said Iverson Scarlett, System Engineer, TAC. “The Performance Lab lets our interns and engineers spot unauthorized PLC writes, validate recipe changes, and teach spectral-based threat hunting—all on the same interface. Rocketgraph has become the bridge between OT data realism and rapid cybersecurity insight for the TAC team.”

About Rocketgraph

Rocketgraph enables enterprises and government agencies to discover the hardest-to-find insights without hiring a command center full of rocket scientists. Born out of a high-performance computing project at the Department of Defense, our graph analytics platform allows an analyst to use GenAI to do iterative analysis with the largest, most complicated datasets on the planet and get answers hundreds of times faster than with traditional graph tools. Rocketgraph builds property graphs that scale to hundreds of billions of edges. Our customers build fine-tuned forecasts, detect sophisticated fraud schemes, monitor nefarious activity on the dark web, keep their networks secure, answer their most challenging questions, and discover the impossible. Experience Rocketgraph in action by signing up for a 30 Day Free Trial or scheduling a demo.

Fonte: Business Wire