- First off, I’m really pleased for the Greenplum team. I’ve gotten to know them well, and they’re a friendly, technically ambitious, no-nonsense bunch (Jeff Cohen’s photoshop efforts notwithstanding). Congrats, folks.
- I like Greenplum’s whole-architecture view of database and analytic technology, and I think it makes sense for EMC. Rather than building their reputation on a single technical differentiator, Greenplum has been steadily integrating good ideas and showing how they can work together at scale. This is good engineering, and good for everybody’s understanding of the technical design space. Many one-feature shops spend a lot of money explaining why things can’t be done, except in their narrow solution. Greenplum has largely been showing what can be done: how good ideas from different quarters can come together over a common scalable and extendible architecture. This will be useful to EMC (see the next points).
- Unlike other acquisitions, EMC didn’t have a database business per se before. So there will be a big role for the Greenplum folks to play, and lot of fun to be had technically. EMC is sometimes referred to as “the smallest of the big companies”. So this acquisition substantially augments the technical profile of a major industry player. It will be a game-changer on all sides. It can also move a lot faster since there’s no need to integrate with a legacy of past offerings.
- EMC is apparently building a new “data computing” division around Greenplum. I like that title. It indicates an ambitious agenda, and jibes with my general view of a future of computing in the cloud that is fundamentally data-driven. It fits well with technical trends I’ve posted about before, regarding both data-centric programming and the decreasing reliance on traditional notions of operating systems in the cloud.
- Seeing this play out in the environment of the VCE Acadia alliance (VMware-Cisco-EMC) will be very interesting, since that crew has the luxury — and arguably the requirement — to rethink the computing stack from scratch. It’s virtualized metal+wires+data. What would you build, long term? Big question.
- In recent months, following their involvement in the MAD Skills work and their Chorus collaboration platform, Greenplum began discussions with academics and practitioners outside the company about open-source tools for Big Data, and the ways that people will work together to make sense of data. I have every indication that this will continue with EMC, perhaps with even more urgency. I have high hopes for this positively helping the open-source ecosystem around Big Data. More on this as it progresses.
Time will tell, but this could end up as more than yet another big-dollar “DBMS startup acquisition”, and even more than “EMC takes on Oracle”. It could be part of a broader technical shift toward general-purpose Data-Driven computing, by players with wherewithal and focus, at a time when technology trends have opened the field to new ideas. That would be interesting for everybody.