Show Student Project Accelerating NumPy's Ufuncs With CorePy for Python Software Foundation

• Google Open Source Programs
  • User (self)
    • Sign In
  • Programs
    • GSoC 2009
      • Home
      • List participating Organizations
      • Apply to become a Mentor
      • List all Student Projects

Release 0-6-20091207

Student Name:	Andrew Friedley
Mentor Name:	christopher mueller
Title:	Accelerating NumPy's Ufuncs With CorePy
Abstract:	Today, most x86 processors contain multiple cores and support SSE vector instructions. However, few open source libraries leverage these features. In this project, I propose optimizing NumPy's ufunc operations by using SSE and multi-core execution via the CorePy library.