mersenneforum.org  

Go Back   mersenneforum.org > Great Internet Mersenne Prime Search > Hardware

Reply
 
Thread Tools
Old 2015-12-11, 12:38   #1
wildrabbitt
 
Jul 2014

6778 Posts
Default Watts to Heat Ratio of a PSU

Hi,

I need to know how much of the X amount of Watts a PSU uses convert to heat.

Is this normally given in the specs for a PSU?
wildrabbitt is offline   Reply With Quote
Old 2015-12-11, 12:49   #2
retina
Undefined
 
retina's Avatar
 
"The unspeakable one"
Jun 2006
My evil lair

614010 Posts
Default

Take 100%, subtract the rated efficiency, and multiply by the usage.

E.g. (100% - 80%) * 500W = 100W

But in reality all of the power gets converted to heat anyway.
retina is offline   Reply With Quote
Old 2015-12-11, 14:11   #3
wildrabbitt
 
Jul 2014

44710 Posts
Default

Thanks Retina.

Is that right ? All the watts get converted to heat ?

Even the beep from POST speaker?
wildrabbitt is offline   Reply With Quote
Old 2015-12-11, 14:13   #4
LaurV
Romulan Interpreter
 
LaurV's Avatar
 
Jun 2011
Thailand

5·1,889 Posts
Default

Quote:
Originally Posted by wildrabbitt View Post
I need to know how much of the X amount of Watts a PSU uses convert to heat.
All of it...
All energy converts to heat one way or another, hehe...

Some of it does useful things for us, like pushing the car forward or lighting the room in the night, before converting to heat, but eventually all ends up as heat...
LaurV is offline   Reply With Quote
Old 2015-12-11, 14:15   #5
wildrabbitt
 
Jul 2014

3·149 Posts
Default

The locations of the heat conversion is what I need to take into consideration.
wildrabbitt is offline   Reply With Quote
Old 2015-12-11, 14:41   #6
chalsall
If I May
 
chalsall's Avatar
 
"Chris Halsall"
Sep 2002
Barbados

2·4,787 Posts
Default

Quote:
Originally Posted by wildrabbitt View Post
The locations of the heat conversion is what I need to take into consideration.
Then in the case of the PSU, use the formula that retina gave. For other devices, ideally measure the power consumed. If that's not possible (often it's not directly) then go with the devices' power consumption ratings. As mentioned, it all ends up as heat at the end of the day.

As an aside, understanding the First and Second laws of thermodynamics is a very empowering tool.
chalsall is online now   Reply With Quote
Old 2015-12-11, 15:47   #7
kladner
 
kladner's Avatar
 
"Kieren"
Jul 2011
In My Own Galaxy!

236568 Posts
Default

Efficiency varies, depending on how heavily loaded the PSU is. I think a rule of thumb is that efficiency peaks around 50% load and trails off in the 85-100% range. This amount of this decline may be related to the rated efficiency of the supply.

If a supply has been reviewed by a technically equipped tester, there may be graphs of its power response. If you plan to run it hard all the time, the benefits of Gold or Platinum ratings increase, which is also true as the total power used increases. I consider that I am in the prime zone drawing a bit over 800 watts from a 1200 watt Platinum supply, on a machine which runs near 24/7.
kladner is offline   Reply With Quote
Old 2015-12-11, 16:31   #8
VictordeHolland
 
VictordeHolland's Avatar
 
"Victor de Hollander"
Aug 2011
the Netherlands

100100110002 Posts
Default

80 Plus certification?
https://en.wikipedia.org/wiki/80_Plus

80Plus: 80% efficient between 20%-100% of (max) rated power
80Plus Bronze: 81% (20% load) 85% (50 %load) 81% (100% load)
80Plus Silver..etc.
80Plus Gold...
80Plus Platinum...
80Plus Titanium...

(see the table in the wikipedia article)

Notice that except for Titanium, they don't have efficiency requirements <20% load. An idle computer could very well be below 20% of the rated wattage of the PSU and be less efficient than 80% even though the PSU has a 80Plus certification.

PSUs with better 80Plus certification (>Silver) are usually more expensive for the same rated power. Depending on your usage and electricity costs, the break-even-point for a more efficient PSU can be months to many years.

Lesson from this: use a PSU which suits your needs (rated power and efficiency vs. costs), and don't lay too much emphasis on getting the most efficient PSU.

In the unlikely case that you are a 'save-the-planet' extremist, don't buy a computer at all! Lot's of energy used to delve the (rare) metals and production probably took place in China which heavily utilises coal plants!
VictordeHolland is offline   Reply With Quote
Old 2015-12-11, 21:36   #9
Mark Rose
 
Mark Rose's Avatar
 
"/X\(β€˜-β€˜)/X\"
Jan 2013

1011011100112 Posts
Default

Quote:
Originally Posted by VictordeHolland View Post
PSUs with better 80Plus certification (>Silver) are usually more expensive for the same rated power. Depending on your usage and electricity costs, the break-even-point for a more efficient PSU can be months to many years.

Lesson from this: use a PSU which suits your needs (rated power and efficiency vs. costs), and don't lay too much emphasis on getting the most efficient PSU.
A reasonable rule of thumb is 1 watt of power costs $1/year ($0.09/kWh). Take a system that consumes 600 watts (two GPUs and a CPU), compare the power supply consumed watts at 50% load vs plain 80 Plus, and the US cost for a 750+ watt power supply (from pcpartpicker.com):

Code:
Unrated:          ???,                  $47
80 Plus:          750,                  $90 (base case for savings below)
80 Plus Bronze:   706, save  $50/year,  $60
80 Plus Silver:   682, save  $68/year,  $81
80 Plus Gold:     667, save  $83/year,  $77
80 Plus Platinum: 652, save  $98/year,  $99
80 Plus Titanium: 638, save $112/year, $377
If high loads will be running most of the time, it's a no-brainer to go with an 80 Plus Platinum power supply. It'll pay for the full cost of the unit in a year, or the upgrade price in a few months. That's not counting air conditioning savings.

If your power is really expensive (California, Germany, etc.), then the Titanium might make sense.

Last fiddled with by Mark Rose on 2015-12-11 at 21:39
Mark Rose is offline   Reply With Quote
Old 2015-12-11, 22:13   #10
Madpoo
Serpentine Vermin Jar
 
Madpoo's Avatar
 
Jul 2014

37·89 Posts
Default

Quote:
Originally Posted by LaurV View Post
All of it...
All energy converts to heat one way or another, hehe...

Some of it does useful things for us, like pushing the car forward or lighting the room in the night, before converting to heat, but eventually all ends up as heat...
Ah, entropy. Delightful.
Madpoo is offline   Reply With Quote
Old 2015-12-12, 09:01   #11
xilman
Bamboozled!
 
xilman's Avatar
 
"π’‰Ίπ’ŒŒπ’‡·π’†·π’€­"
May 2003
Down not across

2×72×109 Posts
Default

Quote:
Originally Posted by LaurV View Post
All of it...
All energy converts to heat one way or another, hehe...
Really?

I would argue that some of it is converted to rest mass. Consider "heat" falling into a black hole, whether in the form of hot matter or electromagnetic or gravitational radiation. Assuming Hawking's mechanism is correct, some of that energy will come out in the form of particles such as electrons and neutrinos. What happens after that is very speculative indeed but a good case has been made for the future universe to consist primarily of positronium, equal numbers of neutrinos and anti-neutrinos and extremely low energy photons.

In the short term, of course, your approximation is an extremely good one.
xilman is offline   Reply With Quote
Reply

Thread Tools


Similar Threads
Thread Thread Starter Forum Replies Last Post
Thread to Post Ratio davar55 Forum Feedback 67 2014-12-06 12:45
limit of a ratio of integrals Mr. P-1 Homework Help 11 2011-11-21 08:08
1.2Ghz machine -- only 5 watts uigrad Hardware 2 2009-02-25 18:46
Ratio of Areas davar55 Puzzles 2 2008-02-13 03:20
Heat problems michaf Sierpinski/Riesel Base 5 3 2006-07-04 21:11

All times are UTC. The time now is 13:50.

Wed May 12 13:50:10 UTC 2021 up 34 days, 8:31, 0 users, load averages: 2.61, 2.44, 2.22

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2021, Jelsoft Enterprises Ltd.

This forum has received and complied with 0 (zero) government requests for information.

Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation.
A copy of the license is included in the FAQ.