Embarcadero did a survey of database administrators (DBAs) of large companies, that was released last month and it worth a look. The survey announcement at http://www.embarcadero.com/press-releases/cloud-computing-to-have-the-greatest-impact-on-database-industry-according-to-study and its title reflects a key marketing element of the survey more than the focus of the survey itself (still, the title is indeed attracting the media like http://www.devx.com/architect/Article/46212/1763). The full survey is a PDF you can get at http://www.embarcadero.com/reports/database-trends-survey, again quite an interesting read also for a developer, maybe one building client/server applications.
While most of the questions are on the choice and number of servers used (MS SQL Server being the most widely adopted, Oracle the most critical, and Firebird has a significant 4% presence for large companies). Other questions are tuned for Embarcadero products (like evaluating the need for automating some DBA tasks). A few questions are interesting for their broad perspective, like those focused on the relationship between developers and DBAs.
However, I want to get back to the only question on "future technologies", the one that took the headlines:
What I don't fully understand is that there is a lot of consideration for "Databases in the cloud" (and that does make sense) and very little for the "noSQL" movement (which is a partial contradiction). In a very significant number of cases, the databases in the cloud are actually based on NoSQL (this is true for Amazon and Azure, among others, although both offer both relational and NoSQL alternatives). This is also due to optimization and architectural reasons. Seems quite a few people who responded have heard about these technologies, but have a limited understanding. Maybe a good topic for writing...
PS. This reminds me I have a half-written article (by me and Paolo Rossi) on programming NoSQL databases from Delphi I should probably try to finish.