Apr 6, 2017
Deepgreen DB 16.16 is out. Major improvements include: better optimization of hash tables, ZStandard compression for column store, and a new SAMPLE keyword for SQL SELECT. Deepgreen DB 16.16 performance has been significantly enhanced as a result of these optimization; it is approximately 15% faster than Deepgreen DB 16.15 when running the full suite of TPC-H 10G. [more]
Aug 14, 2016
Vitesse DB 9.5.3.S is available for download; runs TPCH-Q1 up to 8X faster than Postgres 9.5.3. [more]
May 9, 2016
Deepgreen DB 16.05 is out; now with Fast Decimal, JSON type and faster column store using LZ4 compression algorithm. [more]
Mar 19, 2016
PgDay Asia 2016 talk: Data Warehouse in a PG / Greenplum / Vitesse DB Environment. [slides]
March 23, 2015
Announcing the release of Vitesse DB, a PostgreSQL 9.3 derivative optimized for big workload. TPCH Q1 runs up to 180X faster.
Vitesse Data is a start-up based in Bay Area, California, with deep data expertise.
Most relational databases are still optimizing for I/O while the world grapples with powerful machines that come with lots of memory, fast SSD arrays, and hundreds of underutilized cores. It is not uncommon to find data systems on machines with more memory than the total size of the data. Software innovations, such as column-store and compression techniques, have only served to exacerbate the imbalance.
In view of the new landscape, we created Vitesse Engine — a robust data processing kernel designed to unleash the power of modern x86 CPUs. By fastening Vitesse Engine onto open source databases, we formulate supercharged solutions that are many times more performant, and yet, retain 100% compatibility with the open source projects.