September 2013 Summaries

Oracle OpenWorld has announced new in-memory options in Oracle 12c, with a focus on columnar store technology that leverages compression and parallel query processing. However, this technology has become commoditized and is not enough to solve analytics problems due to the growing data volumes. The company is pushing expensive hardware to justify higher licensing fees, diverting attention from the more important trend of running on commodity hardware. In-memory stores eliminate I/O costs, enabling incredible performance on commodity servers where DRAM is just as fast. The shift towards elastic computing and distributed systems is changing the way customers think about IT infrastructure, with technologies like AWS and private cloud offerings becoming increasingly popular.

The 7th Extremely Large Database Conference is being held at Stanford University next week on September 11, focusing on practical solutions for managing and analyzing extremely large data sets. Digital information now makes up over 98% of the world's stored data, with various new data types being captured at increasingly fast speeds. Modern businesses must analyze this massive stream of incoming data to gain real insights and provide true business value. SingleStore will be represented by its CTO Nikita Shamgunov, who will discuss current database trends and give a live demo of the company's in-memory database solution during the conference on September 10th.