jcwang

This database is incredibly fast, outperforming the STL map and coming very close to the speed of the STL unordered_map.

It supports SQL, transactions, backup/restore, and more.

Get the source code and try now. https://github.com/crossdb-org/crossdb

After one year refactor, new CrossDB is born and open sourced. More features will be added.

Source Code https://github.com/crossdb-org/CrossDB

Document https://crossdb.org/

Official STL Hashmap is unordered_map, which is the fastest data container in CPP. This test is to show how close CrossDB performance to STL Hashmap and CrossDB query can be faster than Hashmap in large data set test.

It's in early PoC dev stage, code is mess, will release later.

STL Map is a popular and fast data container in CPP, but the function is limited. CrossDB is a RDBMS with ACID and it's faster than STL Map.

LMDB is a very fast NoSQL Database, then can a relational database be faster than LMDB? Here comes the new CrossDB and the benchmark is not to compete the features but the performance.

You're right, it's designed for C/C++ at beginning and is not SQL oriented. I plan to develop SQL interface and can be used by many languages. Thanks for the support anyway:)

It's a new designed RDBMS Arch based on memory map with optimized HASH as the Primary Index and use Native APIs to access with user Struct efficiently. There're lots of optimization methods used to improve performance.

Thank you very much, I've corrected the typo. 3.31 is the official library on Ubuntu 20.04 server. I'll do test on new version later.

CrossDB is the Fastest Embedded Database in the world. It's a new RDBMS, optimized for maximum performance. Welcome to do benchmark vs any existing open-source or commercial database.

This database is for C/C++ programmers. It supports RDBMS model: DB, Table, Index, Cursor, etc. It supports transaction. It supports table schema auto-upgrade. It supports ISSU. More to be discovered. Please have a try and share with your friends.