Compute Efficiency Debate

Discussions revolve around comparing computing power in terms of cost per operation, power consumption per joule, and utility across devices like CPUs, mobile SoCs, and servers, questioning whether raw speed or efficiency matters more.

➡️ Stable 0.5x Hardware
3,598
Comments
20
Years Active
5
Top Authors
#1049
Topic ID

Activity Over Time

2007
5
2008
17
2009
41
2010
60
2011
57
2012
108
2013
116
2014
111
2015
148
2016
204
2017
197
2018
178
2019
253
2020
326
2021
308
2022
323
2023
412
2024
352
2025
358
2026
24

Keywords

RAM SoCs CPU GP AI ETL DC AC IO MIP power compute cpu computing cpus computing power cent memory useful processing

Sample Comments

isolli Mar 25, 2021 View on HN

True, it's not all about computing cycles :)

porcc May 31, 2022 View on HN

Compare this with CPU operations per second per dollar and it paints a totally different story

Lofkin May 27, 2015 View on HN

Why can't it be used for general computing?

petermcneeley Nov 21, 2023 View on HN

If you only get 5% of the power of the device why not simply use the CPUs?

bartread Apr 28, 2017 View on HN

We're not talking about computing power here, are we?

sovok_x Sep 24, 2020 View on HN

Yeah, they probably just plain economize on compute there.

aswanson May 26, 2017 View on HN

Probably worse on power consumption as well as price per MIP.

otempomores May 21, 2017 View on HN

As long as the man hours invested cost less then a more potent cpu in all devices sold

imp0cat Jul 16, 2024 View on HN

It depends. There is a lot of devices with quite capable cpus that are mostly doing nothing.

damian2000 Jan 17, 2013 View on HN

I don't think they're targeting people who want or need powerful CPUs