Vectorwise vs OLAP

Question

Would it be preferable to use the Vectorwise based system for efficient data analysis as contrast to OLAP cubes. I have the idea that vector-wise single instruction, multiple data (SIMD) could be good but don't have much good material and resources to design the system.

How can I create a data warehouse in PostgreSQL and then use the vectorwise technology to enhance the fetching, drilling and slicing, dicing operation.

I tried to study the redshift, zoho and they are utilizing/relying on http://www.actian.com/ for their services. What are the good ways to incorporate the cost effective vectorwise technologies.

Answer 1

Looking into different solution and database management techniques, i found that multidimensional arrays are much suitable structure for computing large volumes of data. Vectorwise , HP vertica all are good commercially viable solutions and have the ability of MPP (Massively parallel processing) and SIMD that make them of great use.

I found some open source and multidimensional arrays based DBMS and its comparison can be found here. http://www.slideshare.net/jdegoes/rise-of-the-scientific-database

Answer 2

OLAP can be done in various ways. As you are probably aware of, there are two way of scaling - horisontal and vertical. VectorWise is probably best vertically scalable RDBMS and as such is perfect candidate for OLAP in such environment(s). It is also good (and often used) in collaboration with Hadoop or similar as perhaps the best way of storing large amounts of data for SQL querying.

Finally, the title is misleading and awkward... VectorWise is made for OLAP !! So, there is place for the "vs" in the title.