King's College London Financial Mathematics Centre for Financial Grid Computing

Centre for Financial Grid Computing home page

Publicly Viewable Resources

2010 LGS Parallel Programming Introductions

The following Lectures are extracted from my London Graduate School course notes and might be of wider interest. They are available as Mathematica 7 notebooks or PDFs

1. GridMathematica introduction.

In fact this might be useful to anyone with a standard Mathematica 7, as most of the parallel ideas work with a standard Mathematica 7 supporting 4 slave kernels. Notebook, PDF.

2. Basic Introduction to CUDA

This is a very basic example based overview using a simple vector computation on both a CPU and GPU. Available as Notebook ( I import some timings into Mathematica to analyze) and a PDF.

3. Warps and Threads in CUDA

This is a second basic exampl using a simple vector computation applying quantile functions on both a CPU and GPU. Available as Notebook and a PDF.

These lectures had some Mac-specific makefiles and timing code here and there, but otherwise should be portable. You can find my guide installing GPUs on Mac Pros here.

GridMathematica and IT material

Material below here in Mathematica are Mathematica 6 notebooks unless stated otherwise.

1. Setting up an Apple Mac Pro GridMathematica cluster: Notebook.

2. Presentation at IMS2008 on Monte Carlo simulation and other financial mathematics: Notebook.

2. Presentations at the KCL-Wolfram October 2008 seminar on Financial Grid Computing:

(a) William Shaw on parallel financial applications: Notebook.

(b) Conrad Wolfram (check back)

Workshop on GPU computing sponsored by NVIDIA

This workshop was held in King's in November 2008 and was organized by Professor Claudio Albanese. The official full web page for the course is here. The key presentations are mirrored here as follows:

1. C. Albanese [Level 3 Finance, KCL] on GPU computing [PDF]

2. G. Ziegler [NVIDIA] HPC Computing with CUDA and Tesla hardware [PDF]

3. C. Albanese on Operator Methods for Financial Modelling with GPUs [PDF]

4. C. Albanese on Calibration of an Interest Rate Model Using GPUs [PDF]

5. D. Sayers and A. Jones [NAG], Nag's Libraries and Multi-core for Quant Finance [PDF]

6. M. Giles [Oxford] Financial Computing on NVIDIA GPUs.[PDF]

See also the paper "Eco-mputational Finance" below for a candidate GPU-optimized normal quantile.

KCL Financial Maths Computational Papers/Preprints

Steinbrecher, G. and Shaw, W.T., 2008, Quantile Mechanics, European Journal of Applied Mathematics Vol 19 (2), pp 87-112. doi:10.1017/S0956792508007341: Journal Link.

W.T. Shaw, 2009. Eco-mputational Finance: differential equations for Monte Carlo recycling. Working paper. arXiv:0901.0638v1 [q-fin.CP]