SciDB provides all-in-one data management and analytics. Mix and match data management and analytics in a single query.
Standalone statistical and technical computing software packages like SAS, MATLAB and R don’t store data, they just crunch it. This means that you need to extract data from where it is stored, reformat it into data frames or distributed arrays and then export it to the math package.
These software packages were originally designed to run in-memory only, limiting the amount of data that can be processed. Some have grafted on support to allow them to scale beyond a single computer, but that usually requires significant programming effort by the user. Even if they have partnered with a database vendor or have been integrated by a database vendor, your analytics performance still incurs the internal cost to reformat your data.
Because SciDB queries and scripts scale up with increasing data volumes auto-magically, it can serve as both your analytical workbench and your production solution. Prototype models and scripts can be deployed as-is for production use without recoding.