Use Cases

The canonical use case for standard GridServer Services is risk analysis. Risk is highly computational, with typical run times in the order of seconds or minutes, and it is infinitely hungry for compute nodes, so horizontal scalability is preferred over low latency.

In contrast, a good example of a suitable SpeedLink Service is a Forex trading application, where portfolios, trades, and strategies need to be constantly re-evaluated based on market conditions delivered from a Forex market data feed. This use requires high throughput since Forex rates might change dozens of times a second, and there are hundreds of positions being evaluated. At the same time, the computational requirements for individual positions or strategies are low, so low latency is preferred over horizontal scalability.