Exceedingly Fast Array Operations
SciDB executes array data operations 10–100x faster than a relational database. Conventional databases store data in tables, which are inefficient for large-scale mathematical operations. SciDB’s native array data model delivers dramatic efficiencies in data storage, data access, and computation.
MPP Performance on Commodity Hardware
SciDB’s massively parallel processing engine automatically scales up and accelerates query execution, including complex math operations, as the number of commodity hardware nodes is increased. No need for proprietary appliances or expensive high performance computers to boost your system’s performance as data volumes grow.
Array Data Model
|Data is stored natively as multi-dimensional arrays, well-matched for data with spatial or temporal ordering. High performance math functions run directly on this format.|
Sparse and Dense Arrays
|SciDB’s unique storage management handles both sparse and dense data, resulting in high-performance math, fast access, and compact storage.|
Filtering, Window Operations, and Clustering
|SciDB’s unique storage engine supports fast multi-dimensional filtering, clustering, segmentation, and neighborhood operations.|
Analyze More, Program LessA Webinar about Using SciDB for Computational Finance. Read more.
About Paradigm4Paradigm4 is the company behind the open source SciDB project. We develop it, support it, build enterprise extensions, and provide hands-on expertise.