• Martin Thoma
  • Home
  • Categories
  • Tags
  • Archives
  • Support me

GPUs - Supercomputers for your home

Contents

  • See also:

A few days ago I got some of my neural net code to work with a GPU. The GPU is called "Tesla C2075". It is able to get 515 GFlops peak performance. It has 448 CUDA cores that work with 1.15 GHz and it has 6GB GDDR5 memory.

My code needed about 10 hours to run before. After that, it only needed 10 minutes. That is 60 times faster! The library that did this miracle for me is called Theano)

Out of curiosity, I've searched for current high-end gamer graphic cards. I found nVidia Titan Z:

The Titan Z has 5760 CUDA cores. It can get 4061 GFLOPS x2 and has 12 GB of memory. That technological wonder-work costs only 2802 Euro.

To put that into perspective: In 2005, you would probably have been on place 68 of the TOP500 supercomputers world wide! (source).

Isn't that crazy?

See also:

  • Titan Z Specification
  • What are the differences between “scientific GPUs” and “gaming GPUs”?

Published

Aug 20, 2014
by Martin Thoma

Category

Machine Learning

Tags

  • AI 13
  • CUDA 1
  • GPU 1
  • Machine Learning 81
  • nVidea 1
  • Python 141
  • Theano 1

Contact

  • Martin Thoma - A blog about Code, the Web and Cyberculture
  • E-mail subscription
  • RSS-Feed
  • Privacy/Datenschutzerklärung
  • Impressum
  • Powered by Pelican. Theme: Elegant by Talha Mansoor