Featured Customer: JP Morgan Chase

How GemFire manages exploding data volumes

Intro

Exploding data volumes have become a multi-industry problem. The Financial Services industry in particular has felt the impact. Their business driver has always been to combine historic financial data with real-time, market, trading and valuation/hedging data to get a complete picture of risk so as to make strategic trading decisions. The faster you can do this, the more money you can make. Clearly, this situation presents a massive data management problem both in handling myriads of data and running analysis in sub-millisecond time. GemFire has become the leading data fabric on Wall Street by effectively managing these issues. To show how we do it, we’ll look at the JPMC Pyramid architecture GemFire is an integral part of.

20x speed-up with GemFire

Pyramid is a next generation pricing and risk computation system for equity derivatives trading. The system enables the business to see up-to-date risk numbers throughout the day via a “continuous revaluation” process. The valuations are performed on the JPMCCompute Back-Bone, a compute grid consisting of hundreds of commodity computers fed by GemFire’s data fabric. All of the input data for the models are accessed from the compute engines via GemFire. This new architecture has resulted in a 20x speed-up in processing.