State-of-the-Art.

Backed by a decade of database research.

Icon

A Decade of Research

CedarDB builds on Umbra, a database research project a decade in the making.

Icon

Modern Hardware, First

Designed from the ground up for multi-core CPUs, fast SSDs, and cloud storage.

Icon

Proven Performance

Top scores in ClickBench, TPC-H, and the CH-Benchmark, published and peer-reviewed.

Icon

Built by the Inventors

The same team behind HyPer and Umbra, now building CedarDB.

Research

From Research Project to Production Database

Foundation #1

The Problem

Legacy database systems accumulate 40 years of architectural compromises, leaving them unable to exploit modern hardware.

Our Approach

Umbra was built from scratch to incorporate the lessons of four decades of database research into a system designed for today's hardware.

Foundation #2

The Problem

Most systems force a trade-off between transactional and analytical performance.

Our Approach

Our hybrid storage engine, Colibri, combines compressed columnar and row-based storage so a single system excels at both.

Foundation #3

The Problem

Academic prototypes rarely survive contact with production workloads.

Our Approach

CedarDB turns a decade of peer-reviewed research into an enterprise-grade, PostgreSQL-compatible database you can run today.