VectorWise is a spin-off from the renowned CWI institute in Amsterdam.
Its founders bring a long history of scientific innovation, demonstrated with the MonetDB project.
VectorWise, while incorporating many ideas from MonetDB, is a completely
new database engine, designed from scratch with the goal of combining the
best features of traditional database architectures and MonetDB
and overcome limitations of both.
Our technology is based on the research performed at the CWI database group under the
codename MonetDB/X100.
The architecture of the X100 system is described in the
Data Engineering Bulletin 2005 article.
It proposes numerous technological innovations
improving the performance of
database engines.
The novel vectorized in-cache model of query execution (see the
CIDR 2005 publication)
effectively eliminates many of the
overheads found in the traditional solutions
and manages to fully utilize the performance potential of modern CPUs,
achieving dramatic performance improvement for in-memory processing.
Two techniques are used to scale this performance to disk-resident
data sets.
The concept of lightweight in-cache compression
(see the
ICDE 2006 article)
allows trading some
processor time for an increased data delivery bandwidth.
Additionally, cooperative scans
(see the
VLDB 2007 publication)
dynamically schedule query activity and I/O operations to amortize the cost
of disk access among multiple queries.
As a result, X100 has been shown to provide very high performance both for
in-memory and disk-based database tasks, as well as in
large-scale information retrieval scenarios
(see our
VLDB Journal 2008 article).
Additionally, this technology became a foundation for various
hardware-conscious database techniques, presented in a series
of DAMON papers
(2006,
2007,
and
2008).
VectorWise continues performing cutting-edge research in collaboration with CWI and other academic institutions.