DescriptionJPMorgan Chase runs a large, global business that serves tens of millions of households and 80% of the Fortune 500. Historically, Wall Street has turned to HPC solutions for low-latency trading. In this talk, we describe another domain in which JP Morgan has begun to turn to HPC technology--graph pattern match. Due to the nature of graph challenges, typical cluster-based solutions will not scale to meet our needs. We describe our work to look for known-bad cyber patterns as an application of graph pattern matching on a shared-memory supercomputer.