1 / 4

Use of Python as a MATLAB Replacement for Algorithm Development and Execution in a Multi-Core Environment

Use of Python as a MATLAB Replacement for Algorithm Development and Execution in a Multi-Core Environment. Glen W. Mabey, Ph.D. Southwest Research Institute Signal Exploitation and Geolocation Division San Antonio, Texas. Brian Granger, Ph.D. Tech-X Corporation Boulder, Colorado

clancy
Download Presentation

Use of Python as a MATLAB Replacement for Algorithm Development and Execution in a Multi-Core Environment

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Use of Python as a MATLAB Replacement for Algorithm Development and Execution in a Multi-Core Environment Glen W. Mabey, Ph.D. Southwest Research Institute Signal Exploitation and Geolocation Division San Antonio, Texas Brian Granger, Ph.D. Tech-X Corporation Boulder, Colorado (creator of IPython1)

  2. Requirements • Signals Intelligence algorithm development • Non-cooperative energy detection • Large data sets (½ terabyte) • Signal processing operations • Need to implement algorithms quickly yet be able to run them on large data sets • Distribute over a large number of nodes

  3. Approach • Python: a powerful scripting language • Numpy: numerical arrays for Python • Matplotlib: MATLAB-like plotting facilities • IPython1: • High-level parallel/distributed computing in Python • Supports many styles of parallelism • Fully asynchronous model with asynchronous exception propagation over the network

  4. Results • Input I/O bottleneck addressed using NFS • Output I/O logs to database but also returns intermediate results to client • Wideband recordings processed on 10-dual-dual Xeon blades (40 cores) • 2,653 lines of code • Up to 30x speedup

More Related