Case Study: Numerix achieves 180x faster HPC analytics with Amazon Web Services

Numerix Scales HPC Workloads for Price and Risk Modeling Using AWS Batch

NumeriX, a financial technology company serving more than 250 global clients, needed a better way to scale its HPC-based price and risk modeling as client portfolios grew larger and more complex. Its on-premises approach required costly, ever-expanding hardware, and running detailed analytics across thousands of market scenarios was becoming slow and unwieldy. NumeriX turned to Amazon Web Services, using AWS Batch as part of its cloud migration.

Amazon Web Services implemented a serverless, dynamic batching approach on AWS Batch with Amazon EC2, AWS Step Functions, and Amazon S3 to provision compute only when needed and shut it down after jobs finished. This reduced bottlenecks, eliminated the need for constant hardware expansion, and improved analytics performance by 180 times. As a result, NumeriX can now deliver near-real-time risk analytics, meet 40-minute SLAs on portfolios with tens of thousands of trades, and help clients respond to scaling needs in hours instead of months.

NumeriX

Bill Humphrey

Chief Technology Officer