<?xml version="1.0" encoding="UTF-8"?><rss
version="2.0"
xmlns:content="http://purl.org/rss/1.0/modules/content/"
xmlns:dc="http://purl.org/dc/elements/1.1/"
xmlns:atom="http://www.w3.org/2005/Atom"
xmlns:sy="http://purl.org/rss/1.0/modules/syndication/"
> <channel><title>Comments on: Linux tip: Save Power With Gigabit Ethernet By Using Lower Speed</title> <atom:link href="http://www.cyberciti.biz/tips/linux-ethernet-card-power-saving.html/feed" rel="self" type="application/rss+xml" /><link>http://www.cyberciti.biz/tips/linux-ethernet-card-power-saving.html</link> <description>This is a Linux sys admin journal by Vivek about sys admin work, Linux tips &#38; tricks, hacks, news and more.</description> <lastBuildDate>Fri, 10 Feb 2012 20:37:43 +0000</lastBuildDate> <sy:updatePeriod>hourly</sy:updatePeriod> <sy:updateFrequency>1</sy:updateFrequency> <generator>http://wordpress.org/?v=3.3.1</generator> <item><title>By: Mark Britten</title><link>http://www.cyberciti.biz/tips/linux-ethernet-card-power-saving.html#comment-153118</link> <dc:creator>Mark Britten</dc:creator> <pubDate>Mon, 25 Jan 2010 11:39:21 +0000</pubDate> <guid
isPermaLink="false">http://www.cyberciti.biz/tips/linux-ethernet-card-power-saving.html#comment-153118</guid> <description>this post is quite old, so probably people have moved on from it.
anyway i thought i would chuck my Twocents on the power issue.
If the ethernet connections all operate at the same voltage level, then the slower speeds will save power per byte transmitted. however usually the higher connection speeds are achieved at lower voltages, so the power used to transmit and recieve is lower. moslty it&#039;s not lower than the slower speed per bit but not as bad as it could be.
if a connection carried a signal at 10Mbit/sec it would use say 1unit of power. Now in order to achieve the same volume of data transmission at 1Gbit/sec, the time would be 100x smaller. however the amount of power required would be 100 times greater. this is probably reduced to 30 times the power by adjusting the voltage for transmission and improving the cable quality, but still this means that any power supply will have to produce that much more power to keep all the hardware in the computer alive.
if you have the time to wait for data, it pays for you to do just that.</description> <content:encoded><![CDATA[<p>this post is quite old, so probably people have moved on from it.</p><p>anyway i thought i would chuck my Twocents on the power issue.</p><p>If the ethernet connections all operate at the same voltage level, then the slower speeds will save power per byte transmitted. however usually the higher connection speeds are achieved at lower voltages, so the power used to transmit and recieve is lower. moslty it&#8217;s not lower than the slower speed per bit but not as bad as it could be.</p><p>if a connection carried a signal at 10Mbit/sec it would use say 1unit of power. Now in order to achieve the same volume of data transmission at 1Gbit/sec, the time would be 100x smaller. however the amount of power required would be 100 times greater. this is probably reduced to 30 times the power by adjusting the voltage for transmission and improving the cable quality, but still this means that any power supply will have to produce that much more power to keep all the hardware in the computer alive.</p><p>if you have the time to wait for data, it pays for you to do just that.</p> ]]></content:encoded> </item> <item><title>By: Mark</title><link>http://www.cyberciti.biz/tips/linux-ethernet-card-power-saving.html#comment-145306</link> <dc:creator>Mark</dc:creator> <pubDate>Mon, 20 Oct 2008 04:38:40 +0000</pubDate> <guid
isPermaLink="false">http://www.cyberciti.biz/tips/linux-ethernet-card-power-saving.html#comment-145306</guid> <description>Rather than switching off autonegotation, which can cause problems with mismatched settings, it is better just to limit what your network card announces it is capable of. The other end will autonegotiate up to the fast speed / duplex combination that your end announces. So if you only announce 10 Mbps / Half Duplex, that&#039;s all the other end will use e.g.
# ethtool eth0
Settings for eth0:
Supported ports: [ TP ]
Supported link modes:   10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
1000baseT/Full
Supports auto-negotiation: Yes
Advertised link modes:  10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
1000baseT/Full
Advertised auto-negotiation: Yes
Speed: 100Mb/s
Duplex: Full
Port: Twisted Pair
PHYAD: 1
Transceiver: internal
Auto-negotiation: on
Supports Wake-on: pumbag
Wake-on: d
Current message level: 0x00000001 (1)
Link detected: yes
#
# ethtool -s eth0 advertise 0x001
# ethtool eth0
Settings for eth0:
Supported ports: [ TP ]
Supported link modes:   10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
1000baseT/Full
Supports auto-negotiation: Yes
Advertised link modes:  10baseT/Half
Advertised auto-negotiation: Yes
Speed: 10Mb/s
Duplex: Half
Port: Twisted Pair
PHYAD: 1
Transceiver: internal
Auto-negotiation: on
Supports Wake-on: pumbag
Wake-on: d
Current message level: 0x00000001 (1)
Link detected: yes
#</description> <content:encoded><![CDATA[<p>Rather than switching off autonegotation, which can cause problems with mismatched settings, it is better just to limit what your network card announces it is capable of. The other end will autonegotiate up to the fast speed / duplex combination that your end announces. So if you only announce 10 Mbps / Half Duplex, that&#8217;s all the other end will use e.g.</p><p># ethtool eth0<br
/> Settings for eth0:<br
/> Supported ports: [ TP ]<br
/> Supported link modes:   10baseT/Half 10baseT/Full<br
/> 100baseT/Half 100baseT/Full<br
/> 1000baseT/Full<br
/> Supports auto-negotiation: Yes<br
/> Advertised link modes:  10baseT/Half 10baseT/Full<br
/> 100baseT/Half 100baseT/Full<br
/> 1000baseT/Full<br
/> Advertised auto-negotiation: Yes<br
/> Speed: 100Mb/s<br
/> Duplex: Full<br
/> Port: Twisted Pair<br
/> PHYAD: 1<br
/> Transceiver: internal<br
/> Auto-negotiation: on<br
/> Supports Wake-on: pumbag<br
/> Wake-on: d<br
/> Current message level: 0&#215;00000001 (1)<br
/> Link detected: yes<br
/> #</p><p># ethtool -s eth0 advertise 0&#215;001<br
/> # ethtool eth0<br
/> Settings for eth0:<br
/> Supported ports: [ TP ]<br
/> Supported link modes:   10baseT/Half 10baseT/Full<br
/> 100baseT/Half 100baseT/Full<br
/> 1000baseT/Full<br
/> Supports auto-negotiation: Yes<br
/> Advertised link modes:  10baseT/Half<br
/> Advertised auto-negotiation: Yes<br
/> Speed: 10Mb/s<br
/> Duplex: Half<br
/> Port: Twisted Pair<br
/> PHYAD: 1<br
/> Transceiver: internal<br
/> Auto-negotiation: on<br
/> Supports Wake-on: pumbag<br
/> Wake-on: d<br
/> Current message level: 0&#215;00000001 (1)<br
/> Link detected: yes<br
/> #</p> ]]></content:encoded> </item> <item><title>By: me</title><link>http://www.cyberciti.biz/tips/linux-ethernet-card-power-saving.html#comment-143023</link> <dc:creator>me</dc:creator> <pubDate>Tue, 26 Feb 2008 17:32:53 +0000</pubDate> <guid
isPermaLink="false">http://www.cyberciti.biz/tips/linux-ethernet-card-power-saving.html#comment-143023</guid> <description>how about slower speeds yet like 10 Mbs</description> <content:encoded><![CDATA[<p>how about slower speeds yet like 10 Mbs</p> ]]></content:encoded> </item> <item><title>By: me</title><link>http://www.cyberciti.biz/tips/linux-ethernet-card-power-saving.html#comment-143022</link> <dc:creator>me</dc:creator> <pubDate>Tue, 26 Feb 2008 17:32:22 +0000</pubDate> <guid
isPermaLink="false">http://www.cyberciti.biz/tips/linux-ethernet-card-power-saving.html#comment-143022</guid> <description>how about slower speeds yet, will they save more power</description> <content:encoded><![CDATA[<p>how about slower speeds yet, will they save more power</p> ]]></content:encoded> </item> <item><title>By: No no no</title><link>http://www.cyberciti.biz/tips/linux-ethernet-card-power-saving.html#comment-142952</link> <dc:creator>No no no</dc:creator> <pubDate>Wed, 20 Feb 2008 20:52:11 +0000</pubDate> <guid
isPermaLink="false">http://www.cyberciti.biz/tips/linux-ethernet-card-power-saving.html#comment-142952</guid> <description>No no no babbit.
All grass is green.
All cows eat grass.
Therefore, all cows eat *green* grass.
You&#039;re neither a farmer or an analyst are you?</description> <content:encoded><![CDATA[<p>No no no babbit.</p><p>All grass is green.<br
/> All cows eat grass.<br
/> Therefore, all cows eat *green* grass.</p><p>You&#8217;re neither a farmer or an analyst are you?</p> ]]></content:encoded> </item> <item><title>By: Babbit</title><link>http://www.cyberciti.biz/tips/linux-ethernet-card-power-saving.html#comment-142457</link> <dc:creator>Babbit</dc:creator> <pubDate>Sat, 12 Jan 2008 23:27:41 +0000</pubDate> <guid
isPermaLink="false">http://www.cyberciti.biz/tips/linux-ethernet-card-power-saving.html#comment-142457</guid> <description>All grass is green.
All cows eat grass.
Therefore, all cows are green.</description> <content:encoded><![CDATA[<p>All grass is green.<br
/> All cows eat grass.<br
/> Therefore, all cows are green.</p> ]]></content:encoded> </item> <item><title>By: Neil</title><link>http://www.cyberciti.biz/tips/linux-ethernet-card-power-saving.html#comment-141287</link> <dc:creator>Neil</dc:creator> <pubDate>Fri, 05 Oct 2007 17:03:04 +0000</pubDate> <guid
isPermaLink="false">http://www.cyberciti.biz/tips/linux-ethernet-card-power-saving.html#comment-141287</guid> <description>I have no opinion on the power saving, but a new GigE switch that we got didn&#039;t want to play nice with a bunch of our GigE-capable boxes.  I was trying to figure out how to tell the NICs to run at 100Mbps, and this article showed me how.  Much appreciated.</description> <content:encoded><![CDATA[<p>I have no opinion on the power saving, but a new GigE switch that we got didn&#8217;t want to play nice with a bunch of our GigE-capable boxes.  I was trying to figure out how to tell the NICs to run at 100Mbps, and this article showed me how.  Much appreciated.</p> ]]></content:encoded> </item> <item><title>By: DasIT</title><link>http://www.cyberciti.biz/tips/linux-ethernet-card-power-saving.html#comment-141185</link> <dc:creator>DasIT</dc:creator> <pubDate>Mon, 24 Sep 2007 18:53:09 +0000</pubDate> <guid
isPermaLink="false">http://www.cyberciti.biz/tips/linux-ethernet-card-power-saving.html#comment-141185</guid> <description>Efficiency is not just a measure of total wattage.  It is a measure of performance per watt, be it in CPU cycles or data rates.  If I save 1 watt by lowering my link speed to 100Mbps, it now takes me nearly an order of magnitude longer (not exactly an order of magnitude, due to protocol ineffeciencies, but close anyway) to copy the same amount of data of over the same data link that would normally be transmitting at 1Gbps, how is that a savings?  It&#039;s actually costing me more time and therefore more wattage over time to copy the same amount of data.  If anything, people should go out and install 1Gbps or even 10Gbps on all of their equipment to save as much as they can!
I sincerely appreciate that people are trying to save energy, but energy savings at the cost of output is a false savings.  It&#039;s not about being lazy or impatient, its about making the most of your investments, both in grid power and man power.</description> <content:encoded><![CDATA[<p>Efficiency is not just a measure of total wattage.  It is a measure of performance per watt, be it in CPU cycles or data rates.  If I save 1 watt by lowering my link speed to 100Mbps, it now takes me nearly an order of magnitude longer (not exactly an order of magnitude, due to protocol ineffeciencies, but close anyway) to copy the same amount of data of over the same data link that would normally be transmitting at 1Gbps, how is that a savings?  It&#8217;s actually costing me more time and therefore more wattage over time to copy the same amount of data.  If anything, people should go out and install 1Gbps or even 10Gbps on all of their equipment to save as much as they can!</p><p>I sincerely appreciate that people are trying to save energy, but energy savings at the cost of output is a false savings.  It&#8217;s not about being lazy or impatient, its about making the most of your investments, both in grid power and man power.</p> ]]></content:encoded> </item> <item><title>By: Jacky</title><link>http://www.cyberciti.biz/tips/linux-ethernet-card-power-saving.html#comment-141172</link> <dc:creator>Jacky</dc:creator> <pubDate>Sat, 22 Sep 2007 18:51:29 +0000</pubDate> <guid
isPermaLink="false">http://www.cyberciti.biz/tips/linux-ethernet-card-power-saving.html#comment-141172</guid> <description>&lt;blockquote cite=&quot;VonSkippy&quot;&gt;So if I had 1000’s of servers it might be worth the effort to make the change - it’s certainly not worth it for 47 servers&lt;/blockquote&gt;
This is exactly the type of thinking we have too much of on this world, simple minded.
0.5W times every 1000Mbit computer out there is A LOT of WATTS saved every day. Get of your lazy ass TODAY, save some power and allow you children to enjoy a better world 500 years from now ... it&#039;s about the big picture.....</description> <content:encoded><![CDATA[<blockquote
cite="VonSkippy"><p>So if I had 1000’s of servers it might be worth the effort to make the change &#8211; it’s certainly not worth it for 47 servers</p></blockquote><p>This is exactly the type of thinking we have too much of on this world, simple minded.<br
/> 0.5W times every 1000Mbit computer out there is A LOT of WATTS saved every day. Get of your lazy ass TODAY, save some power and allow you children to enjoy a better world 500 years from now &#8230; it&#8217;s about the big picture&#8230;..</p> ]]></content:encoded> </item> <item><title>By: VonSkippy</title><link>http://www.cyberciti.biz/tips/linux-ethernet-card-power-saving.html#comment-141162</link> <dc:creator>VonSkippy</dc:creator> <pubDate>Fri, 21 Sep 2007 21:02:11 +0000</pubDate> <guid
isPermaLink="false">http://www.cyberciti.biz/tips/linux-ethernet-card-power-saving.html#comment-141162</guid> <description>Where did you get the 2 Watt per system savings figure?  I tested several of my servers, mostly Supermicro MB&#039;s or IBM&#039;s (using the Kill-A-Watt realtime power monitor) and saw almost no change (&gt;0.5 watt) when switching between 1000 and 100 ethernet speeds.
So if I had 1000&#039;s of servers it might be worth the effort to make the change - it&#039;s certainly not worth it for 47 servers.</description> <content:encoded><![CDATA[<p>Where did you get the 2 Watt per system savings figure?  I tested several of my servers, mostly Supermicro MB&#8217;s or IBM&#8217;s (using the Kill-A-Watt realtime power monitor) and saw almost no change (&gt;0.5 watt) when switching between 1000 and 100 ethernet speeds.</p><p>So if I had 1000&#8242;s of servers it might be worth the effort to make the change &#8211; it&#8217;s certainly not worth it for 47 servers.</p> ]]></content:encoded> </item> </channel> </rss>
