TE
TechEcho
Home24h TopNewestBestAskShowJobs
GitHubTwitter
Home

TechEcho

A tech news platform built with Next.js, providing global tech news and discussions.

GitHubTwitter

Home

HomeNewestBestAskShowJobs

Resources

HackerNews APIOriginal HackerNewsNext.js

© 2025 TechEcho. All rights reserved.

Race for embedded AI chips is on

2 pointsby greendeskover 8 years ago

1 comment

throwaway387463over 8 years ago
&gt;As a research organization, CEA has been studying how best to bring deep neural networks to edge computing. Asked about barriers to DNN on edge computing, Duranton said it’s clear that “floating point” server solutions cannot be applied, because of “power, size and latency constraints.” Other limitations include: “a number of MACs, bandwidth and on-chip memory size,” he added.<p>&gt;OK. “Integer instead of floating” is one of the first things we need to consider.<p>I wonder why they don&#x27;t try to implement Unums[1].<p>[1] - <a href="https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Unum_(number_format)" rel="nofollow">https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Unum_(number_format)</a>