NVIDIA Grace Hopper Superchip Boosts Murex MX.3 Analytics, Cuts Power Usage

NVIDIA’s Grace Hopper Superchip is making significant strides in enhancing the performance of Murex’s MX.3 analytics platform, according to an announcement on the NVIDIA Blog. Murex, a renowned trading and risk management software company, has adopted this advanced technology to better manage credit and market risk exposures in derivatives contracts.

Managing Risk with MX.3 Driven by Grace Hopper

Financial institutions require robust computing infrastructure to run complex risk models on vast datasets and deliver real-time decision-making capabilities. The Murex MX.3 platform, used by over 60,000 daily users across 65 countries, provides comprehensive coverage for credit and market risk, BASEL capital standards, and x-valuation adjustments (XVA).

Murex is currently testing the Grace Hopper Superchip for various calculations, including XVA, market risk calibration, pricing evaluation, sensitivity, and profit and loss assessments across different asset classes. The Grace Hopper Superchip not only enhances calculation speed but also offers significant power savings.

“On counterparty credit risk workloads such as CVA, Grace Hopper is the perfect fit, leveraging a heterogeneous architecture with a unique mix of CPU and GPU computations,” said Pierre Spatz, head of quantitative research at Murex. “On risk calculations, Grace is not only the fastest processor but also far more power-efficient, making green IT a reality in the trading world.”

Initial tests in the Murex research and development lab show that Grace Hopper can reduce energy consumption by 4x and improve performance by 7x compared to traditional CPU-based systems.

Pricing FX Barrier Options in MX.3 with Grace Hopper

In addition to risk management, Murex is leveraging the Grace Hopper Superchip to price foreign exchange (FX) barrier options using its latest stochastic local volatility model. Barrier options are derivatives whose payoff depends on whether the underlying asset price reaches or crosses a specified threshold during the contract period. The pricing evaluation, which involves solving a 2D partial differential equation, is more cost-effective on the Arm-based NVIDIA Grace CPU in GH200.

Running these calculations on the Grace Hopper Superchip is 2.3x faster compared to the Intel Xeon Gold 6148 processor. Additionally, the NVIDIA Grace CPU offers significant power efficiencies, being 5x better on a watts-per-server basis for FX barrier calculations.

NVIDIA’s next-generation accelerated computing platform is proving to be a game-changer for high-performance computing in quantitative analytics for the capital markets. The impressive results in terms of performance and energy efficiency position the Grace Hopper Superchip as a crucial asset for financial institutions looking to optimize their operations.