EMC hasit will acquire Greenplum, a data warehousing and business analytics software firm for an undisclosed sum. EMC will use this acquisition to form the basis of a new Data Computing Products Division led by Bill Cook, CEO of Greenplum, who will report to Pat Gelsinger, COO of EMC's Information Infrastructure Products. To put that statement into perspective, Backup and Recovery Solutions (where Data Domain and other related acquisitions now live) is also a separate EMC division reporting to Gelsinger. BRS is a big division with a lot of products. Therefore, I think one can safely bet that Data Computing Products at EMC will grow in scale and scope.
And here we see that EMC's marketing minds were hard at work. While EMC is positioning Greenplum in business analytics, this new division is not being called the EMC Business Analytics Division, nor the Data Warehousing/Business Intelligence Division, or the even sexier Cloud Analytics Division. No. This is the Data Computing Products Division. What is data computing or a data computing product? I'll let EMC explain because I'm not sure that I can.
First observation: A common assumption is that EMC is doing this to respond to Oracle's success with its Exadata solution and NetApp's acquisition of Bycast. True, EMC's Data Computing solutions yet to be seen will likely compete with the new Sun Oracle as systems vendor. But we should add to the list Hewlett-Packard, IBM, Teradata, database vendors, cloud software vendors, and anyone else in the new, cloud friendly, business analytics space. EMC sees Greenplum's database and cloud portal technologies as disruptive to the traditional data warehousing/business intelligence market. And, in the possession of EMC's worldwide marketing and sales force, Greenplum could very well be disruptive--within the DW/BI marketplace and beyond. Read the specs on Greenplum's massively parallel processing (MPP) database product. They're impressive. What's even better for EMC is that the Greenplum Database likes to be integrated with storage. This acquisition cries out for an integrated hardware software stack (think VBlock) or a yet-to-be-named EMC Data Computing appliance (GBlock?). Is EMC ooching toward becoming a systems vendor? No. Not without a major services group.
Right now, EMC just wants to be a major player in the fast-growing business analytics segment--hence the Greenplum acquisition and the creation of the EMC Data Computing Products division. However, the challenge for EMC will be to grow a presence in a space where its large and formidable sales force is virtually unknown. In EMC's favor is the fact that DW/BI is rapidly evolving from a relatively slow batch processing application that pulls data from a few sources, to one that takes massive amounts of data from a variety of sources including real time sensory data, and delivers results in real or near-real time to many concurrent users. Business analytics is a new opportunity for all.
Second observation: We've seen very little mention made so far about Greenplum Chorus which is even more germane to EMC's "Journey to the Private Cloud" strategy. Greenplum calls it the "first commercial Enterprise Data Cloud." I'm not here to argue the validity of that statement. I only wish to point out that anyone who attended EMC World 2010 couldn't have helped but notice the Journey to the Private Cloud signage. It was everywhere. Greenplum Chorus does cloud-based self-service provisioning of data marts, allows cloud analytics users to share data sets and data marts with others, and supports social networking and collaboration. As such, Chorus fits right in as a major journey-to-the-cloud destination. Let me restate my previous conjecture in a slightly different way: cloud data analytics is a new opportunity for all.
Expect to see a few of Greenplum's partners defect. (HP and Sun are good bets here, I think.) However, integration projects with open-source software (Hadoop for example) will proceed forward. Expect to see some more EMC Data Computing products relatively soon like an EMC/Greenplum data computing appliance or at least some DIY reference models. In a recent post, EMC blogger extraordinaire Chuck Hollis alludes to running parts of the Greenplum processing stack on any storage array that uses x86 processing technology. That allusion could presage a new optimized array from EMC or a version of a current EMC storage array that supports Greenplum Database processing offload.
EMC's Data Computing Products division is a Pat Gelsinger creation. Can he and Bill Cook be disruptive with it? Only time will tell.