Big Data is everywhere. Predictive analytics and real time in-memory computing isn’t everywhere.
This truth (if we can accept it to be so) represents something of an imbalance.
As a subset of data mining, predictive analytics driven by in-memory computing efficiencies now has an opportunity to bring real-time analysis and insight to fast-moving live transactional data flows. Or to put it another (rather shorter) way, we can now start to manage and understand Big Data better than ever.
If we combine contemporary approaches to predictive analytics with the newly arrived Intel Xeon Phi coprocessor that produces what is claimed to be over one teraflop per second in terms of workload computational power for highly-parallel workloads, then CIOs can start to think about what “50-processor core computing” will mean for us in the very near future.