With the general availability of its SAP HANA appliance in June, the bigger picture is starting to come into focus, so we thought it was time to try to set the record straight on some HANA naming conventions, many of which have changed recently.

SAP has been using a number of different names for various pieces of HANA technology but, with June 20^th General Availability, some of them have gone out the window.  The one name you will be hearing quite a bit is simply “SAP HANA,” which SAP defines as a platform for real-time analytics and applications. But wait – HANA is not an acronym or abbreviation any longer, so it doesn’t get further spelled out, according to a blog post by Amit Sinha, VP of Technology and Innovation Platform Marketing at SAP, with responsibility for in-memory computing & SAP HANA.

Sinha details these name changes:

• SAP High-Performance Analytic Appliance is now just SAP HANA appliance or SAP HANA application cloud, depending on deployment model
• SAP In-Memory Appliance is now SAP HANA
• SAP In-Memory Computing Engine (SAP ICE or IMCE) is now SAP HANA database (this also replaces any previous usage of “newDB”)

Defining SAP HANA

While the naming convention is simple enough, defining exactly what HANA is and does is a bit more complicated.

While it was initially envisioned as an analytics engine that would simply help customers perform queries much faster because they were conducted in memory, HANA is now more than that. “SAP HANA evolved very quickly from a high-performance analytical appliance to a ‘platform’ for real-time analytics AND applications,” according to Sinha.

Still, the HANA appliance is at the core of everything; the trick is getting it to connect to whatever resources you want to speed up, and to some sort of client that enables the whole thing to be useful to end users.  That requires either a direct interface to HANA or some sort of intermediary layer.

Out of the gate, according to an SAP webinar on HANA featuring Uddhav Gupta, Sr. Solution Manager for Enterprise Data Warehouse platforms, clients including SAP BusinessObjects Explorer, SBO Analysis for MS Office and Microsoft Excel will be able to directly connect to the SAP HANA appliance. Others, such as Crystal Reports, Web Intelligence and SAP BusinessObjects Dashboards, will require use of a semantic layer to perform some translations.

By the fourth quarter of 2011, SAP NetWeaver Business Warehouse will be able to run completely on HANA database – but only if you’re using the latest version, 7.30, with SPS 5.  Still, at that point you’ll be able to take your existing BW system and replace the underlying database with HANA, Gupta said in the webinar. “Nothing else has to change,” he said. That will put all your BW data into HANA and push many calculations from the OLAP engine into HANA.

SAP NetWeaver BW is one of the many applications which SAP envisions will ultimately run natively on SAP HANA. In December of 2010, SAP announced the first application built on HANA, Strategic Workforce Planning. Other apps scheduled for this year include:

• Sales and operations planning
• Cash and liquidity management
• Trade promotions management
• Smart meter analysis
• Customer revenue performance management
• Customer-specific pricing
• Profitability engine
• Merchandising and assortment management
• Energy management for utility customers

At SAPPHIRE Now 2011, SAP customer Centrica demoed its use case for Smart Meter Analytics application running on HANA. Smart meter analysis is one example of future out-of-the-box apps “Powered by HANA” that could fuel the HANA market by providing an instant use case for analytical apps that can process records at speeds superior to classic OLAP databases.

Early customer results

Customers who participated in the HANA ramp-up must have access to these resources, however, and they’re reporting impressive results, according to Gupta.

A large bank migrated a database into HANA and saw an immediate compression ratio of more than 6:1, with no schema changes.  That led to a query speed improvement of 369 times.  And that was with no optimization, Gupta said.

In another test for a customer, SAP took data from 70 retailers, 50 terabytes of it, and put it on HANA.  It saw a compression rate of almost 10:1 and, using Explorer to conduct queries, saw a response time of 0.04 seconds – even though the test involved searching through 460 billion records.

Another interesting story will be the addition of in-memory components to other SAP applications. For example, in November, SAP Business One will begin ramp-up on a side-by-side HANA appliance working alongside the B1 instance to provide enhanced speed and report querying options, with further B1-HANA integration to come in 2012. SAP Business ByDesign already runs its analytics in-memory, with further in-memory capabilities to follow. Now that the naming conventions are pretty well sorted out, HANA integration across SAP applications and the creation of new HANA-based apps will be the story to watch.

Editor’s note: thanks to SAP Mentor Vitaliy Rudnytskiy for his contributions to this article. For more details on HANA, check out John Appleby’s SAP HANA FAQ and SAP’s in-memory section on SCN.

The ERP Executive is published by Panaya.