Home Embedded Software Multimedia Enabling High Performance Computing on Embedded Systems

Enabling High Performance Computing on Embedded Systems

On-demand Web Seminar

Requires Flash Player.

High Performance Computing or HPC has classically been the domain of super computers consisting of rooms filled with endless rows of rack-mounted processor blades. However, recent advances in processors and single board computers are enabling compute-intensive applications to run on embedded systems. This session reviews the challenges of accessing the teraflop-range performance potential of these new platforms and explains how high level libraries like Sourcery VSIPL++ and performance profiling tools like System Analyzer can help maximize application speed performance and portability while ensuring high developer productivity.

Duration: 31:51

Products: Sourcery VSIPL++

View On-demand Web Seminar (Opens in New Window/External URL)

Details

Overview

High Performance Computing (HPC) has classically been the domain of super computers consisting of rooms filled with endless rows of rack-mounted processor blades. However, recent advances in processors and single board computers are enabling compute-intensive applications to run on embedded systems. This session reviews the challenges of accessing the teraflop-range performance potential of these new platforms and explains how high level libraries like Sourcery VSIPL++ and performance profiling tools like System Analyzer can help maximize application speed performance and portability while ensuring high developer productivity.

What You Will Learn

Common pitfalls to accessing maximum hardware speed performance
The advantages of using a high-level library for HPC application development
How performance profiling data visualization can accelerate HPC application debugging

About the Presenters

Brooks Moses

Dr. Moses leads the High Performance Computing Solutions team in Mentor Graphics' Embedded Software Division. He also participates directly in the development of the Sourcery VSIPL++ library and other high-performance library products. Dr. Moses worked extensively on the Cell/B.E and NVIDIA CUDA ports of Sourcery VSIPL++. Dr. Moses holds a Ph.D. in Mechanical Engineering from Stanford University where he conducted advanced research into algorithms for computational fluid dynamics simulation.

Pete Decher

Pete is the Business Development Manager for High Performance Computing Solutions at Mentor’s Embedded Systems Division. Pete has over 30 years of experience in software development, electronic system design and test at a variety companies ranging from startups to Fortune 500. Pete holds a BSEE degree from Georgia Tech and an MSEE degree from Stanford University as well as several US patents.

Who Should View

Software engineers developing compute-intensive embedded applications
Engineering managers focusing on software development cost reduction and process improvement
Software engineers tasked with maintenance and platform migration of compute-intensive applications

Related Resources

Multimedia

How to Measure RTOS Performance

This session takes a look at "typical" reported performance metrics for an RTOS in the embedded industry. It's an attempt to explain what these numbers signify, how they are measured, and why...…View On-demand Web Seminar

Creating & Debugging Linux Apps Using CodeBench for ARM

Learn how to use Sourcery CodeBench to develop and debug embedded Linux applications. This video shows how to create an application for a target that is running Linux, download the application to the target,...…View Technology Overview

Technology and Tips for Developing Rich UIs on IPTV and STB Products

Consumers continue to find user interfaces on TV products, including Set-Top Box and IPTV, to fall short of expectations in a technology rich world where products have rich, dynamic UIs. Many product manufacturers...…View On-demand Web Seminar