Making a Case for In-Memory Database

Dynamic random access memory is getting cheaper, and
ever growing set of applications is becoming fully RAM-resident.

In this talk I'll state the case for in-memory technology in purely engineering terms: how memory-focused algorithms
and data structures create a performance and efficiency edge over traditional systems, significant enough to
justify an own product family:

- taking performance to the next level: why only an in-memory database memory manager can reach the levels of speed and efficiency unattainable by general purpose memory allocators

- bringing concurrency to the data, and why single-core concurrency can surpass multi-core parallelism in performance

- in-memory database bottlenecks: what are they, and how they are different from a traditional DBMS.