TIBCO introduces ComputeDB following its acquisition of SnappyData
TIBCO describes the transition of SnappyData technology into ComputeDB.
Press & Media
Selected coverage, official stories, interviews, and public announcements.
Read announcement ↗
Contention-aware transaction scheduling arrives in InnoDB
MySQL credits contributed University of Michigan research behind contention-aware scheduling in InnoDB.
Barzan Mozafari named Morris Wellman Faculty Development Professor
A University profile of research contributions, open-source systems, and technology translation.
U of Michigan Project Combines Modeling and Machine Learning
HPCwire covered the ConFlux project, combining high-performance computing, large-scale data infrastructure, and machine learning for physical modeling.
A case for predictable databases
An interview with Barzan Mozafari about predictable database performance, DBSeer, DBSherlock, and translating systems research into open-source tools.
Simplifying Interactive, Real-Time And Advanced Analytics
Forbes highlighted BlinkDB among systems advancing interactive and approximate analytics over large datasets.
Interactive Big Data analysis using approximate answers
O'Reilly described BlinkDB's sampling-based approach to interactive analysis, including explicit tradeoffs between response time and answer accuracy.
New MIT Software Targets Data-Intensive Cloud Computing
Inside Big Data highlighted DBSeer as an open-source system for modeling database workloads and improving data-intensive cloud operation.
MIT Researchers Pushing for a Better-Run Cloud
The Wall Street Journal's CIO Journal covered DBSeer and its approach to modeling database workloads for more efficient cloud resource management.
Software seer shows companies path to cheaper databases
The Register covered DBSeer's use of workload modeling to predict database resource requirements and reduce cloud operating costs.
Can't Hire Big Data Staff? Try Enterprise Crowdsourcing
InformationWeek covered research using active learning to reduce the cost of crowdsourcing labels for very large datasets.
BlinkDB: a query engine for making perfect decisions in the absence of perfect answers
HP Labs profiled BlinkDB's use of approximate answers and statistical error guarantees for interactive queries over massive datasets.