High performance computing

Grid computing

  The pricing of derivatives with complex diffusion models such as local or stochastic volatility usually require the use of a CPU grid.

With a strong knowledge of financial software as well as derivative products and models, we provide efficient answers to performance matters:

- grid deployment
- grid configuration
- model adaptation on grid
- validation of grid solutions.

GPU for computational finance

  The performance gains offered by GPUs are incomparable to those achievable with other technologies available today. In that domain, Monte-Carlo simulations are the obvious beneficiaries of GPU technology.

This method is of course heavily used in exotics pricing - path-dependent products, multi-asset payoffs, models with more than 3 factors - as well as in the risk field (value-at-risk, potential future exposure...). Its adaptation to GPUs allows an incredible gain in speed, ranging from 60 times faster computations.

Although the adaptation of partial differential equation (PDE) methods to GPU is more complex, it can also benefit largely from GPU technology. In that domain, Nvidia’s latest Tesla 20-series Fermi presents particularly promising capabilities.

With a solid knowledge of financial models and computation, our specialists can guide your organisation on the high performance track.

  Copyright 2009 Upskills Intranet | Webmail | Disclaimer  



Upskills becomes a Murex preferred Partner and certified integrator in the Asia Pacific region.

Hot jobs

- Murex Front Office Consultant

- Senior Project Manager
(Abu Dhabi)

- Front Office Developer


November 2011 (Singapore)
The Battle of the Quants full day event enables quantitative focused hedge funds, investors and technology providers to meet, discuss and listen to cutting edge innovations and latest trends in the quantitative field of finance. This event, now in its fourth year, comes to Singapore for the first time in November 2011.


Upskills is a Murex partner.

Murex partner