I believe in the balance between diversity and homogeneity. Homogeneity helps the society and individual groups to achieve the goal efficiently. Diversity helps the society and individual groups to select and refine the goal efficiently. The following contemporary interests of mine reflect my effort of balancing diversity and homogeneity grown out of my past years in academia.
Next-Generation In-Memory Database Engine
- To exploit hundreds of gigabytes of memory and ever-increasing processing power of modern multi-core computer architecture, we invented a new generation of DBMS kernel algorithms such as parallel differential logging and recovery, and cache-conscious index structures and concurrency control. P*TIME, a full-fledged DBMS that materialize these breakthroughs, was first developed with the funding from Transact In Memory, Inc., an SNU laboratory venture established in Silicon Valley aiming the global market by Prof. Cha, who then led it to a successful strategic merger with SAP AG, the global market leader in business software, in late 2005. Today, the research in this area, continues in collaboration with SAP AG, leveraging the dedicated R&D infrastructure of tens of 64GB/128GB servers at SAP R&D Center Korea.
Massively Parallel Cloud Data Management
- Google has pioneered the development of a massively parallel and distributed search service infrastructure on top of tens of thousands of inexpensive commodity hardware boxes. As such a massive scale of hardware infrastructure becomes available as commercial cloud computing service, we investigate architectural issues for providing real-time data management service on the cloud with a wide range of application domains in mind, from ubiquitous intelligence to scientific discovery. This investigation includes exploring architectural alternatives of massively scalable and space-efficient storage management weighing between row and column stores, new ways of managing replicas for fault tolerance, and devising query processing strategies for the new architecture.
Teaching young engineers with entrepreneurial mind
- Engineering education needs to go further beyond teaching how to solve a given technical problem. Entrepreneurial mind involves recognizing the important problem to be solved, recognizing the value of the technology, and planning the execution of rolling out the new technology to the market.