cmu-db/noisepage

Health v2 · maintenance only

Flags: archived

Adoption not part of the score

1765 stars · 501 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

NoisePage is a relational database management system from Carnegie Mellon University designed for autonomous, self-driving operation, with integrated machine learning components that forecast, model, and plan system behavior. It offers Postgres wire-protocol compatibility, lock-free MVCC, LLVM-based JIT query compilation, and Arrow-compatible columnar storage.

Use cases

• self-driving autonomous database management system
• postgres-compatible database with machine learning tuning
• research DBMS with JIT query compilation
• lock-free multi-version concurrency control database
• experiment with ML-driven database optimization
• columnar in-memory storage database for research

When to choose

• you are researching autonomous database optimization and self-tuning systems
• you need a Postgres-protocol-compatible experimental DBMS with JIT compilation
• you want to study or extend a modern MVCC columnar storage engine
• you are a CMU student or contributor working on database internals

When to avoid

• you need a production database with broad platform support and long-term vendor support
• you require environments other than Ubuntu 20.04, which is the only officially supported build target
• you need a stable, actively released database - the latest release dates to 2022
• you just need a simple embedded or lightweight database for an application

Facets

application · maturity maintenance

database machine-learning llm-training databases machine-learning cpp self-hosted dbms self-driving-database postgres-compatible research-project llvm jit-compilation mvcc apache-arrow research linux

Member repositories

For agents

markdown · JSON · MCP: product_card(name="cmu-db/noisepage")