Data Warehouse

Speedminer Data Warehouse (DW) supports advanced features to facilitate secure and extremely efficient data transfer, especially when handling data loading from remote sites. The Speedminer DW Agent accesses data from remote sites and performs the checking to extract only changed data elements and queue them as data packets to be transferred to the DW Server.  The data packet transfer is extremely secure and even if the connection is dropped, the system shall continue monitoring until the connection is up again, and then continue to transfer the data across from where it left off. Thus, this minimizes the bandwidth requirements and achieves maximum performance in building the Enterprise Data Warehouse.

 

Traditional Business Intelligence is built upon rigid online analytical processing (OLAP) architecture which limits ability due to many restrictions.  Among them are the restriction of adaptability to changes (due to complex design for Star Schema, Dimension Table and Fact Table), restriction of having only a limited number of dimensions (more dimensions means slower in performance), restriction on performance (especially when dealing with arbitrary time span and multi-pass queries) and many other restrictions. Although some Business Intelligence (BI) providers claim to have solutions to the restrictions, there is a catch as other restrictions are imposed, such as requiring excessively large amounts of memory and also high specifications on hardware.

Speedminer Data Warehouse engine is a native multidimensional data engine supporting all RDBMS features, Object Database features and Multi-Value feature.  This is the third manifesto data storage technology that unify the relational and object in native manner, thus this is the best platform for any multi-dimensional analysis.

FEATURES

The single version of truth stored in a powerful environment support SQL, Object Query, as well as the most user friendly querying environment – Implicit Join with SQL Shortcut.  This feature only available in Speedminer has proven to enable knowledge building on query to allow even non-technical person with no SQL background to be able to generate powerful, complex and accurate easily, of course with the implementation of Business Intelligence, querying data can be even faster and more powerful.

Speedminer Memory Management Technology uses efficient memory processing to deliver superb performance to OLAP queries. The memory is segregated to 8K, 16K, 32K and 64K memory blocks that efficiently stores and processes the information in the memory to reduce the need for retrieving information from the hard disk. Unlike conventional in-memory technologies which require huge amounts of memory to host all the information, Speedminer Memory Management is designed to store most frequent accessed information in the memory, thus producing the best performance based on the available hardware. Speedminer is the only OLAP engine with the support for multiple types of memory blocks. This ensures maximum performance for different data size requirements.

Speedminer employs an OLAP technology that supports the Third Normal Form (Cubeless OLAP) which surpasses all other OLAP technologies and does not impose the impediments and limits of traditional OLAP architecture as mentioned above. Many database vendors and BI vendors hope to eventually have a unified OLAP and OLTP environment.  Speedminer is the only BI that has unified OLAP and OLTP. What is the significance of unified OLAP and OLTP?

 

An OLAP environment is capable of answering complex queries at a faster pace as compared to an OLTP environment. But an OLAP environment is not capable of supporting Third Normal Form, meaning that OLTP architecture support is absent. This results in the inability to navigate through different data elements (records) according to the database relationship. Navigation of data elements involves browsing between related information. For example, from a purchase record, users can navigate to the items purchased, → goods received of the item, → vendor supplying the item, → price list or other related information.

 

Without the need for cumbersome Cubes or Fact tables, the implementation of data warehouse has also become simple and fast.

Speedminer OLAP uses superior technology to convert data and store them as compressed bitmap in columnar format to enable analysis of huge data within split seconds.  With these technologies, Speedminer does not only deliver faster performance, it also requires less storage and delivers consistent performance as compared to other OLAP technologies.