The Data Computing Appliance, first deliverable from EMC’s acquisition of Greenplum, was announced last month, only 75 days after the acquisition closed, and it doesn’t lack for ambition. Pat Gelsinger, President and Chief Operating Officer, EMC Information Infrastructure, pointed to the high level opportunity: unlocking the “hidden value” of enormous and growing data assets every company is increasingly holding, and often failing to leverage. The appliance will reach many hitherto untapped resources in the data centers that EMC occupies. Adding EMC’s manufacturing, sales and marketing, and reference architectures to the Greenplum IP brings what Gelsinger calls Greenplum’s “first phase” to its completion. And begins what is likely to be a sizable battle with Oracle, Teradata and IBM, if EMC mounts campaigns and spending to match its ambitious vision.
The Data Computing Appliance has a preloaded copy of Greenplum Database version 4.0 on a full or partial rack or racks of servers, storage and networking in an MPP configuration. A single rack has 16 servers, with two Intel E5670 6-core, 2.93GHz processors each. Each rack holds up to 36 terabytes of uncompressed storage, and 24 of them (with 4,608 cores) can be combined to hold 5 petabytes after compression across 24 racks. All this is Dell hardware today, but one can imagine we will see other hardware opportunities ahead, especially given that Greenplum will run on EMC’s Vblock infrastructure packages from the Virtual Computing Environment coalition. Note that Cisco plays there – and wants to get into the hardware business too. I’d imagine other shoes will drop before too much time goes by. Meanwhile, EMC is calling itself the biggest and baddest box on the planet. Click the image to enlarge.
Some of the key marketing themes are already evident in the early story: from the label (The Answer Machine) to the attributes EMC is leading with: performance, volume, load time and “private cloud virtualized infrastructure for analytics.” EMC has tapped into the key themes of the emerging market for analytic platforms and is ready to talk the talk. Rolling out customers like the operator of the NY Stock Exchange (NYSE Euronext) and ClickFox to drive the message home, EMC can throw considerable weight behind an offering that had achieved substantial momentum before its acquisition.
Two key expertises drive the story: EMC’s management of massive data volumes (NYSE Euronext takes in transaction records amounting to terabytes per day; ClickFox analyzes half a billion customer interactions a month, also in TB volume) and Greenplum’s ability to derive analytic value from it to grow business opportunity. EMC’s Data Domain deduplicated backup and recovery software can be added for DW backup to EMC (or non-EMC) SANs; RecoverPoint software can also be used to add replication for additional data center-class hardening.
Speed of load drives speed of delivery. Co-founder Scott Yara says this is the “fastest data loading machine on the planet” – it will be fun to see the sparring among the players. In the launch videos, ClickFox alludes to its need to build models in real time, and cites speed as one of the reasons it migrated from Oracle, whose 12-hour plus processing of a key query it has reduced to 10 minutes. EMC’s willingness to poke the folks in Redwood Shores is significant. Although Greenplum could perhaps have been largely ignored as it picked off 140 customers in its first couple of years, EMC’s entry in the market will certainly be more of a threat to Oracle. And strategically important to EMC – given Oracle’s aggressive push into storage sales with its bundling of Exadata storage into its principal market offering, a key battleground has been signaled.
This is a fundamental point for the ADBMS market – its growth does not need to come at the expense of the existing market share of the leading RDBMSs; rather, the battle is at the margin, for new opportunities. The whole pie will grow for a while, and this is what has permitted a thousand customers to opt for new ADBMS platforms – $350 million worth in 2009, even without Teradata and Sybase IQ – without any evident slowdown for Oracle or IBM in their growth rates. (For more on this, see the report Colin White and I completed this summer.)
Luke Lonergan, Greenplum’s co-founder, sounded EMC’s major theme of value: “insight is the key competitive asset today” as companies look for advantages in their data, he said. It’s about analytics. The Greenplum team needs to be sure to keep that story near the top of the chart and not be overwhelmed by the natural inclination of EMC to push speeds and feeds, enterprise-class management of data as an asset, and the like. These are competitive necessities at the high end, but what got ADBMSs here, and permitted even the little players like Greenplum to grow rapidly, was bringing analytics to the data. Greenplum must continue to showcase its analytic capabilities, and counter the thrusts of firms like Aster Data, ParAccel and Vertica, who are pushing that side of the story, even while it takes on the big guns with its added data center attributes.
Disclosure: Greenplum is a client of IT Market Strategy