carli2

MemCP targets to be a drop-in replacement for existing row-based databases and offers a performant column-based alternative that exceeds reacable speed limits of row-based databases. It targets to users who want performance, users who don't want to license a professional database like ORACLE or SAP HANA but want the professional performance.

We just launched MemCP, a blazing-fast, open-source, columnar in-memory database designed to handle both OLAP and OLTP workloads. Built in Go, MemCP combines cutting-edge parallelization, spaltenbasierte Speicherung (columnar storage), and intelligent compression (1:5 average) to deliver performance 10x faster than traditional databases like MySQL — with 80% less storage usage.

Features:

    In-memory design for sub-millisecond query speeds.
    Columnar storage for efficient analytics and transactional workloads.
    Modern hardware optimizations: NUMA, NVMe, multicore CPUs.
    Lightweight and portable—embed directly into your app!
    Built-in REST API support for rapid app development.

Why it matters: MemCP offers the power of enterprise-grade analytics at open-source accessibility. Whether you’re running a startup or scaling a cloud-native app, MemCP is here to level up your data game.

Learn more & contribute on GitHub Early adopters welcome—feedback drives us forward!

Discussion prompt: Can MemCP disrupt MySQL/MariaDB in the same way Redis shook up caching? What killer apps could this enable?

We at launix are MySQL power users and crash our MySQL servers on a daily basis. Therefore we implemented our own SQL database protocol compatible to MySQL which offers the following features: - 10x faster performance for aggregation queries - in-Memory storage - in-Memory compression reduces RAM requirements by 80% - in-database microservices where microservice code and SQL queries are precompiled for best performance