Tutorial 6: Why Programming Many-Core is Not Mission Impossible
Monday, June 3, 2013
Time: 11:00 AM — 1:00 PM
|Organizer: ||Michael Solka - Coherent Logix, Austin, TX|
|Speakers: ||Bryan Schleck - Coherent Logix, Austin, TX|
| ||MK Sandeep - Coherent Logix, Austin, TX|
Attendees will learn how a many-core processor can be programmed using C and the industry standard Message Passing Interface "MPI". A DSP processor based on HyperX™ technology will be used as the demonstration vehicle. A structured approach to system analysis will be described that produces an optimal mapping of an algorithm to processing resources.
Mr. Bryan Schleck
An example will be used to show in-depth analysis and subsequent mapping to processing resources. Additionally, a full LTE PHY and an H.264 encoder will be demonstrated. Throughout the tutorial, an interactive Eclipse-based environment is used for development and running the resulting programs on a cycle-accurate instruction set simulator and target hardware.
The intended audience for this tutorial is algorithm developers of embedded systems. Knowledge of C programming is required. Familiarity with basic DSP principles will be helpful. Concepts covered will include parallel programming methods, structured system and algorithm analysis, and embedded system development and debug.
holds an M.S. in electrical engineering from the University of Minnesota and a B.A. in engineering physics from St. Mary’s University. Mr. Schleck’s responsibilities at Coherent Logix include training and support for customers who are using the many-core HyperX processor and algorithm development using C and MPI on the HyperX. In the past, he has led the design and simulation of algorithms on the HyperX processor in various algorithm applications demonstrating critical performance metrics. These algorithms include parallel implementations of FFT, correlation, convolution, motion estimation, feature tracking algorithms, hyperspectral imaging, space based fusion, and Synthetic Aperture Radar.
Mavuduru Kannappa is a DSP Applications Engineer at Coherent Logix. His responsibilities include system design and implementation of software-defined radio and video processing applications on the HyperX many-core processor. He has worked on the design, parallelization, implementation, and integration of different modules in H.264 video encoder, LTE PHY layer, and Digital RF front end algorithms on the HyperX processor.
He holds a M.S in Electrical Engineering from The University of Texas at Dallas and a B.Tech in Electronics and Communications Engineering from National Institute of Technology, Calicut, India.