Hacker GitHub

← All repos

keroro824/HashingDeepLearning

View on GitHub ↗ ★ 1,105 C++

Codebase for "SLIDE : In Defense of Smart Algorithms over Hardware Acceleration for Large-Scale Deep Learning Systems"

5 mentions on Hacker News by 4 people: 0 stories, 5 comments · first 2020-03-04 · latest 2021-04-09

Mentions per year 5 mentions · busiest year 2020 (3)
20082026
AllStoriesComments NewestTopOldest
  1. Paper: https://arxiv.org/abs/1903.03129 Code: https://github.com/keroro824/HashingDeepLearning Edit: Here is the new paper referred to https://arxiv.org/abs/2103.10891 Here is the new code: https://github.com/RUSH-LAB/SLIDE The paper I…
    comment by exegete · 2021-04-09 · in CPU-based algorithm trains deep neural nets up to 15 times faster than top GPU
  2. https://arxiv.org/abs/1903.03129 https://github.com/keroro824/HashingDeepLearning
    comment by exegete · 2021-04-09 · in CPU-based algorithm trains deep neural nets up to 15 times faster than top GPU
  3. ….rice.edu/2020/03/02/deep-learning-rethink-overc... paper: https://arxiv.org/pdf/1903.03129.pdf repo: https://github.com/keroro824/HashingDeepLearning
    comment by cztomsik · 2020-03-07 · in Deep Learning Breakthrough Sees One Xeon Destroying 8x Nvidia Tesla V100 GPUs
  4. Why do you all guess? The paper has a link to their github with the source and instructions: https://github.com/keroro824/HashingDeepLearning I see no reason not to try it with some AMD Threadripper or EPYC instead.
    comment by LargoLasskhyfv · 2020-03-06 · in Smart Algorithms beat Hardware Acceleration for Large-Scale Deep Learning
  5. source: https://github.com/keroro824/HashingDeepLearning paper: https://arxiv.org/abs/1903.03129
    comment by yayr · 2020-03-04 · in Slide: First algorithm for training deep neural nets faster on CPUs than GPUs