Trading desks have always struggled to balance I/O performance with the capacity of data that must be retained in their market data and research archives. To address this, backtest data is often tiered across a pyramid of storage systems that are designed either for fast I/O or large capacity. On the other hand, the abundance of market data that each firm maintains creates added opportunity to find new correlations and to improve the accuracy of trading models. As quants continue to evolve the questions they ask their data – it becomes almost impossible to reprocess or re-analyze data that’s been exiled to slow, archival storage. The problem of storage bottlenecks has become such a contentious issue, that quants are now asking “what storage do you research on” as part of the interview process.