nTersect

GPU Technology Conference Highlights: Finance

Posted: 19 Aug 2010 02:13 PM PDT

We're four weeks away from the GPU Technology Conference (GTC) and during the lead up, we wanted to highlight some of the content you can expect to see at the show. Previously, we told you about great high performance computing sessions, and this week we're focusing on how GPUs play a part in computing for financial markets.

At last year's GTC, we also covered the topic of GPU computing and finance, in fact you can check out some of those archived sessions here:

Overview of Computational Finance on the GPU by NVIDIA
Using GPUs to Estimate the Value-at-Risk of Portfolios by UC Berkeley and UC Davis
GPU Acceleration of Mortgage CMO Option Adjusted Spread by Bloomberg

Computational Finance on the GPU

This year you can see even more great finance sessions, including:

Integrating GPGPU Accelerated Pricing Models into an Existing Financial Services Infrastructure (Session #2033): Scott Donovan, Citadel Investment Group.

Join Citadel Investment Group to explore our three year undertaking on the feasibility of GPGPU computing for option pricing. We will discuss our 140X performance boost and the hurdles we had to overcome to integrate GPUs into our existing infrastructure. Please note that our talk will not get into the details of the model (that's proprietary information), but we will share our innovative solution to drive a grid of GPUs.

Practical Methods Beyond Monte Carlo in Finance (Session #2032): Pierre Spatz, Murex

Murex will share its practical experience using GPUs to accelerate high-performance analytics based on GPU-enabled Monte Carlo and PDE methods. We will also briefly describe Murex's experience developing a high-level payoff scripting language that allows user-definable payoffs for single and cross-asset instruments.

Derivatives & Bond Portfolio Valuation in a Hybrid CPU/GPU Environment (Session # 2040): Peter Decrem, Quantifi

Learn how to compute traditional end of day computations in real time through the use of a hybrid GPU/CPU computing environment. We will detail how computing intensive tasks are delegated to the GPU while interface issues are dealt with by the CPU. We will discuss our methodology consisting of the following three components: (1) valuations; (2) by tenor risk measures; and (3) full distributions allowing for more complex analytics such as exotic options products valuation and counterparty value adjustments calculation.

Attendees can still take advantage of the early bird discount before September 1 by registering here. Once registered you can also set up your personal schedule using the online booking system, guaranteeing your seat in the sessions you want to attend.

Beauty Box: Adobe Photoshop Plug-in for Portrait Retouching

Posted: 19 Aug 2010 11:36 AM PDT

If you've ever spent time retouching portraits and close-ups, you know the challenges of removing an imperfection while maintaining the details of the picture.The latest CUDA accelerated Adobe Photoshop Plug-in, Beauty Box, helps you do just that.

You can smooth wrinkles in skin and even out skin tones by sliding the controls until you're happy with your portrait (if only this worked in real-life). And, since Beauty Box is CUDA accelerated, turning on your NVIDIA GPU increases your PC's responsiveness when previewing your adjustments. It also shortens the time it takes to save your final picture by up to 7 times.

Check out this video to see how Beauty Box helps make portraits picture perfect, and get a trial version of Digital Anarchy's Beauty Box here.