GPU Accelerated framework for financial nested simulations

Joris Cramwinckel
ORTEC Finance
ORTEC Finance, 2015

   title={GPU Accelerated framework for financial nested simulations},

   author={Cramwinckel, Joris},



Download Download (PDF)   View View   Source Source   



In this thesis we present a state-of-the-art approach to accelerate Monte Carlo valuations of embedded options. Due to regulations and improved risk management, nested simulations (scenarios in scenarios) are becoming increasingly important for institutional investors like: insurance companies, pension funds and housing corporations. Preferably one wishes to use a framework in which multiple related problems of nested simulations can be accelerated with GPUs. We build such a framework using advanced CUDA features from NVidias Kepler and higher architectures. CUDA streams and Hyper-Queues enable the GPU to run tasks effectively by running concurrent and overlapping CPU and GPU calculations. In addition, NVidias Multi Processing Service is used to handle offloading Monte Carlo valuations from different local processes to the GPU. Runtimes in the order of days of current implementations are reduced to the order of minutes. This broadens the horizon of the current nested simulation methodologies. Besides, the proposed framework is scales well with the number of compute nodes and GPUs per cluster.
VN:F [1.9.22_1171]
Rating: 5.0/5 (1 vote cast)
GPU Accelerated framework for financial nested simulations, 5.0 out of 5 based on 1 rating

* * *

* * *

TwitterAPIExchange Object
    [oauth_access_token:TwitterAPIExchange:private] => 301967669-yDz6MrfyJFFsH1DVvrw5Xb9phx2d0DSOFuLehBGh
    [oauth_access_token_secret:TwitterAPIExchange:private] => o29ji3VLVmB6jASMqY8G7QZDCrdFmoTvCDNNUlb7s
    [consumer_key:TwitterAPIExchange:private] => TdQb63pho0ak9VevwMWpEgXAE
    [consumer_secret:TwitterAPIExchange:private] => Uq4rWz7nUnH1y6ab6uQ9xMk0KLcDrmckneEMdlq6G5E0jlQCFx
    [postfields:TwitterAPIExchange:private] => 
    [getfield:TwitterAPIExchange:private] => ?cursor=-1&screen_name=hgpu&skip_status=true&include_user_entities=false
    [oauth:protected] => Array
            [oauth_consumer_key] => TdQb63pho0ak9VevwMWpEgXAE
            [oauth_nonce] => 1477649221
            [oauth_signature_method] => HMAC-SHA1
            [oauth_token] => 301967669-yDz6MrfyJFFsH1DVvrw5Xb9phx2d0DSOFuLehBGh
            [oauth_timestamp] => 1477649221
            [oauth_version] => 1.0
            [cursor] => -1
            [screen_name] => hgpu
            [skip_status] => true
            [include_user_entities] => false
            [oauth_signature] => +eOJY8L/QYu6PVS3wtNXlRZEu8U=

    [url] => https://api.twitter.com/1.1/users/show.json
Follow us on Facebook
Follow us on Twitter

HGPU group

2037 peoples are following HGPU @twitter

HGPU group © 2010-2016 hgpu.org

All rights belong to the respective authors

Contact us: