Behind SAP’s Business Intelligence Accelerator

In an article published today, eWeek claims: SAP may have stumbled onto an Oracle killer: in-memory technology that could, in theory, quash the need for a relational database in some cases.

This Rather Incomplete eWeek Article Makes The Following Points:

With Business Intelligence Accelerator (BIA) SAP has sussed out a way to organize its business intelligence data in columns versus tables, storing and indexing the data in memory and then running it all on blade servers.
The result is faster queries than would be possible by tapping data stored in a data warehouse or relational database.
Coca Cola showed a 90 percent increase in reporting performance, with queries cut from 60 seconds down to 3 seconds in the case of Coca-Cola using BIA
SAP is working on in-memory data management capabilities that could go beyond BI to other areas of the application stack, replacing the need for a relational database in new software installations.

Either eWeek Didnt Do its Homework or Just Flat Out Neglected to Mention:

SAP uses the same column-based or vertical decomposition approach for data compression within BI accelerator that has been employed for years by Sand Technologies and Sybase.
SAP customers that want fast queries STILL need ONE BIA server and BIA software licenses for each instance of SAP NetWeaver BI, in addition to the underlying databases. 

So, BIA duplicates data from SAP NW BI into the BIA, adds another piece of hardware, and requires additional (optional) SAP BIA licenses. Net, Net, BIA adds complexity to already complex BW implementations.
SAP Business Intelligence Accelerator is SAPs workaround for poor performance and scalability of SAP NW BI systems.

SAP BW customers plagued by the notoriously poor performance and limited scalability of

SAP BW systems are now held hostage by SAP to purchase BIA as an expensive option to fix BWs performance and scalability problems.

SAP BW is already known to be the most expensive data warehouse to build and maintain (according to recent OLAP Survey). This additional set of costs only adds to the exorbitant TCO of SAP NW BI (and helps SAP recover lost SAP BW revenue streams from bundling BW for free with NetWeaver).
SAP BIA can only leverage SAP BW Info Cubes as a data source. It does not work with other SAP data sources such as an ODS, let alone other heterogeneous data sources. What happens when performance and scale are needed in a heterogeneous environment? 

 No surprise here – a SAP BW implementation in a heterogeneous environment is like Big Foot, we all hear rumors about its existence, but no ones ever seen one.
There is currently no failover mechanism for BI accelerator (i.e., if the system goes down, the indexes will have to be rebuilt from beginning).
SAP BIA is compatible with SAP NW BI 2004s, so customers must first upgrade to even have the opportunity to pay more and complicate their infrastructure in exchange for performance.
Oracle already has a huge leg up when it comes to in memory database technology as a result of its acquisition of leading in memory database vendor, TimesTen
There are no independent benchmarks for BIA to date, only claims from SAP.

Comments are closed.