There seems to be no end to the stream of acquisitions in recent months. First SAP buying Sybase, then HP paid a ridiculous price for 3PAR due to competing bid from Dell. HP followed that within weeks with another more-than-you-should-pay acquisition of ArcSight. EMC bought GreenPlum not too long ago. In the same space, IBM buys Netezza paying quite a premium ($1.7B).
Why this acquisition of Netezza is significant to IBM? I remember the early days of Netezza when one of the investor/board members spoke to me. Jeet Saxena was the founder of this Boston-based company that wanted to bring a turn-key appliance for high-speed data warehousing and analytics. So the goal was to go “thin and deep” as opposed to “wide and thin” as offered by the traditional commercial RDBMS products or even the “cubes”. So Netezza was the first one (after Teradata) to bring scale-out parallel processing for high volume data reduction and analysis – bundling hardware and software together as an appliance. They built their software pieces on IBM hardware. hence it was natural for IBM to buy them. GreenPlum did the same thing on Sun hardware, but after Oracle bought Sun, it needed a new home. Hence EMC acquiring GreenPlum made sense, as Oracle’s core strength is database, data warehousing, and parallel processing. I still do not think Oracle has a true Netezza-like data warehousing appliance, notwithstanding the marketing of Exadata. We will wait to see how well Exadata (now Exalogic elastic computing cloud) competes with a IBM-Netezza combo solution. You may think of IBM offering Netezza as a specialized “warehousing engine” separate from its DB2 warehousing solutions.
Netezza is more modern in its architecture. It rode the wave of these trends – excessive amount of data collection, don’t-throw-away-any-data mentality, get some advantage out of these massive data collections, and cheaper processing and storage costs. It has been a strong alternative to the more expensive Teradata type solutions of the past in terms of speed and cost.
However, IBM needs to explain its positioning to the customers more clearly – when to use DB2 for warehousing and analytics vs. Netezza. Besides DB2 and Netezza, IBM also has Informix and the former Redbricks, also Cognos analytics stuff. But IBM is used to the syndrome of “many products to solve the same problem”.
After this acquisition, I have read somewhere speculations of Oracle looking at Vertica, another column-based-search company from Michael Stonebraker. I won’t be surprised if that happens. It is a horse race now!