<?xml version='1.0' encoding='UTF-8'?><?xml-stylesheet href="http://www.blogger.com/styles/atom.css" type="text/css"?><feed xmlns='http://www.w3.org/2005/Atom' xmlns:openSearch='http://a9.com/-/spec/opensearchrss/1.0/' xmlns:georss='http://www.georss.org/georss' xmlns:gd='http://schemas.google.com/g/2005' xmlns:thr='http://purl.org/syndication/thread/1.0'><id>tag:blogger.com,1999:blog-7364458407847551072</id><updated>2011-07-28T15:08:19.806-04:00</updated><title type='text'>Losing Watts</title><subtitle type='html'>Low-power computing and networking.</subtitle><link rel='http://schemas.google.com/g/2005#feed' type='application/atom+xml' href='http://lpcomp.blogspot.com/feeds/posts/default'/><link rel='self' type='application/atom+xml' href='http://www.blogger.com/feeds/7364458407847551072/posts/default?max-results=100'/><link rel='alternate' type='text/html' href='http://lpcomp.blogspot.com/'/><link rel='hub' href='http://pubsubhubbub.appspot.com/'/><author><name>David Andersen</name><uri>http://www.blogger.com/profile/03996590425188586871</uri><email>noreply@blogger.com</email><gd:image rel='http://schemas.google.com/g/2005#thumbnail' width='31' height='32' src='http://bp0.blogger.com/_Kj9HEENopvM/SIX7dbPIK4I/AAAAAAAAB_E/39s4kYcj_kc/s1600-R/dave1-sm.jpg'/></author><generator version='7.00' uri='http://www.blogger.com'>Blogger</generator><openSearch:totalResults>17</openSearch:totalResults><openSearch:startIndex>1</openSearch:startIndex><openSearch:itemsPerPage>100</openSearch:itemsPerPage><entry><id>tag:blogger.com,1999:blog-7364458407847551072.post-1748249709759792032</id><published>2009-05-01T10:52:00.001-04:00</published><updated>2009-05-01T10:53:43.654-04:00</updated><title type='text'>More folks join the datacenter PUE race:  Fortune Data Centers</title><content type='html'>&lt;a href="http://www.treehugger.com/files/2009/04/fortune-data-centers-starts-up-greener-operations-in-silicon-valley.php"&gt;Using overhead cold air and insulated hot air return ducts&lt;/a&gt; to improve PUE.  Article is light on technical details.&lt;div class="blogger-post-footer"&gt;&lt;img width='1' height='1' src='https://blogger.googleusercontent.com/tracker/7364458407847551072-1748249709759792032?l=lpcomp.blogspot.com' alt='' /&gt;&lt;/div&gt;</content><link rel='replies' type='application/atom+xml' href='http://lpcomp.blogspot.com/feeds/1748249709759792032/comments/default' title='Post Comments'/><link rel='replies' type='text/html' href='http://lpcomp.blogspot.com/2009/05/more-folks-join-datacenter-pue-race.html#comment-form' title='39 Comments'/><link rel='edit' type='application/atom+xml' href='http://www.blogger.com/feeds/7364458407847551072/posts/default/1748249709759792032'/><link rel='self' type='application/atom+xml' href='http://www.blogger.com/feeds/7364458407847551072/posts/default/1748249709759792032'/><link rel='alternate' type='text/html' href='http://lpcomp.blogspot.com/2009/05/more-folks-join-datacenter-pue-race.html' title='More folks join the datacenter PUE race:  Fortune Data Centers'/><author><name>David Andersen</name><uri>http://www.blogger.com/profile/03996590425188586871</uri><email>noreply@blogger.com</email><gd:image rel='http://schemas.google.com/g/2005#thumbnail' width='31' height='32' src='http://bp0.blogger.com/_Kj9HEENopvM/SIX7dbPIK4I/AAAAAAAAB_E/39s4kYcj_kc/s1600-R/dave1-sm.jpg'/></author><thr:total>39</thr:total></entry><entry><id>tag:blogger.com,1999:blog-7364458407847551072.post-5903502831885147356</id><published>2009-04-02T15:04:00.001-04:00</published><updated>2009-04-02T15:07:04.563-04:00</updated><title type='text'>Google's server blade design unveiled</title><content type='html'>&lt;blockquote&gt;Google's big surprise: each server has its own 12-volt battery to supply power if there's a problem with the main source of electricity. The company also revealed for the first time that since 2005, its data centers have been composed of standard shipping containers--each with 1,160 servers and a power consumption that can reach 250 kilowatts&lt;/blockquote&gt;
&lt;p&gt;&lt;a href="http://news.cnet.com/8301-1001_3-10209580-92.html?tag=contentMain;contentBody"&gt;Google uncloaks once-secret server&lt;/a&gt;&lt;/p&gt;
&lt;p&gt;&lt;img src="http://i.i.com.com/cnwk.1d/i/bto/20090401/GoogleServerMedium.jpg"&gt;&lt;/p&gt;&lt;div class="blogger-post-footer"&gt;&lt;img width='1' height='1' src='https://blogger.googleusercontent.com/tracker/7364458407847551072-5903502831885147356?l=lpcomp.blogspot.com' alt='' /&gt;&lt;/div&gt;</content><link rel='replies' type='application/atom+xml' href='http://lpcomp.blogspot.com/feeds/5903502831885147356/comments/default' title='Post Comments'/><link rel='replies' type='text/html' href='http://lpcomp.blogspot.com/2009/04/googles-server-blade-design-unveiled.html#comment-form' title='0 Comments'/><link rel='edit' type='application/atom+xml' href='http://www.blogger.com/feeds/7364458407847551072/posts/default/5903502831885147356'/><link rel='self' type='application/atom+xml' href='http://www.blogger.com/feeds/7364458407847551072/posts/default/5903502831885147356'/><link rel='alternate' type='text/html' href='http://lpcomp.blogspot.com/2009/04/googles-server-blade-design-unveiled.html' title='Google&apos;s server blade design unveiled'/><author><name>David Andersen</name><uri>http://www.blogger.com/profile/03996590425188586871</uri><email>noreply@blogger.com</email><gd:image rel='http://schemas.google.com/g/2005#thumbnail' width='31' height='32' src='http://bp0.blogger.com/_Kj9HEENopvM/SIX7dbPIK4I/AAAAAAAAB_E/39s4kYcj_kc/s1600-R/dave1-sm.jpg'/></author><thr:total>0</thr:total></entry><entry><id>tag:blogger.com,1999:blog-7364458407847551072.post-6364407862975637334</id><published>2009-03-23T21:26:00.002-04:00</published><updated>2009-03-23T21:33:03.918-04:00</updated><title type='text'>HotOS paper list released</title><content type='html'>The &lt;a href="http://www.usenix.org/event/hotos09/tech/tech.html"&gt;list of accepted papers for HotOS 2009&lt;/a&gt; is out, and power starts to show up heavily.  The final papers won't be ready before the end of the semester, most likely, but perhaps we'll be able to get a few pre-copies to look at.  Obviously power-relevant titles include:
&lt;ul&gt;
&lt;li&gt;Operating System Support for NVM+DRAM Hybrid Main Memory (expect to see more of this over the next few years)&lt;/li&gt;
&lt;li&gt;Mobility Changes Everything in Low-Power Wireless Sensornets&lt;/li&gt;
&lt;li&gt;FAWNdamentally Power-efficient Clusters&lt;/li&gt;
&lt;li&gt;&lt;i&gt;perhaps&lt;/i&gt;: Peloton: Coordinated Resource Management for Sensor Networks  (sensor "resources" almost always include power)&lt;/li&gt;
&lt;li&gt;&lt;i&gt;perhaps&lt;/i&gt;: Augmented Smart Phone Applications Through Clone Cloud Execution (title sounds similar to the Kimberly paper we read)&lt;/li&gt;
&lt;/ul&gt;&lt;div class="blogger-post-footer"&gt;&lt;img width='1' height='1' src='https://blogger.googleusercontent.com/tracker/7364458407847551072-6364407862975637334?l=lpcomp.blogspot.com' alt='' /&gt;&lt;/div&gt;</content><link rel='replies' type='application/atom+xml' href='http://lpcomp.blogspot.com/feeds/6364407862975637334/comments/default' title='Post Comments'/><link rel='replies' type='text/html' href='http://lpcomp.blogspot.com/2009/03/hotos-paper-list-released.html#comment-form' title='0 Comments'/><link rel='edit' type='application/atom+xml' href='http://www.blogger.com/feeds/7364458407847551072/posts/default/6364407862975637334'/><link rel='self' type='application/atom+xml' href='http://www.blogger.com/feeds/7364458407847551072/posts/default/6364407862975637334'/><link rel='alternate' type='text/html' href='http://lpcomp.blogspot.com/2009/03/hotos-paper-list-released.html' title='HotOS paper list released'/><author><name>David Andersen</name><uri>http://www.blogger.com/profile/03996590425188586871</uri><email>noreply@blogger.com</email><gd:image rel='http://schemas.google.com/g/2005#thumbnail' width='31' height='32' src='http://bp0.blogger.com/_Kj9HEENopvM/SIX7dbPIK4I/AAAAAAAAB_E/39s4kYcj_kc/s1600-R/dave1-sm.jpg'/></author><thr:total>0</thr:total></entry><entry><id>tag:blogger.com,1999:blog-7364458407847551072.post-3982372365378327800</id><published>2009-03-06T11:20:00.003-05:00</published><updated>2009-03-06T11:33:55.633-05:00</updated><title type='text'>Physical storage</title><content type='html'>In class, we talked about different ways of storing data -- as charges, as magnetic polarization, and using physical or chemical changes.  My jokes about punch cards aside, physical storage does have potential benefits -- particularly in terms of longevity.  Michael De Rosa points out two research projects in this area:
&lt;ul&gt;
&lt;li&gt;&lt;a href="http://en.wikipedia.org/wiki/IBM_Millipede"&gt;IBM's Millipede&lt;/a&gt; uses a MEMS array of probes (much like the magnetic MEMS arrays we talked about), but these probes heat and are used to physically deform the surface of a piece of acrylic glass.&lt;/li&gt;
&lt;li&gt;&lt;a href="http://www.norsam.com/hdrosetta.htm"&gt;Norsam's HD-Rosetta&lt;/a&gt; project etches data into a nickel plate, which can then be read either optically or with an electron microscope.  The product is based upon &lt;a href="www.osti.gov/bridge/servlets/purl/762925-5UlmIO/webviewable/762925.PDF"&gt;research from Los Alamos National Lab&lt;/a&gt;.  They claim that HD-Rosetta can store 23-400 gigabits / sq in.
&lt;/ul&gt;&lt;div class="blogger-post-footer"&gt;&lt;img width='1' height='1' src='https://blogger.googleusercontent.com/tracker/7364458407847551072-3982372365378327800?l=lpcomp.blogspot.com' alt='' /&gt;&lt;/div&gt;</content><link rel='replies' type='application/atom+xml' href='http://lpcomp.blogspot.com/feeds/3982372365378327800/comments/default' title='Post Comments'/><link rel='replies' type='text/html' href='http://lpcomp.blogspot.com/2009/03/physical-storage.html#comment-form' title='0 Comments'/><link rel='edit' type='application/atom+xml' href='http://www.blogger.com/feeds/7364458407847551072/posts/default/3982372365378327800'/><link rel='self' type='application/atom+xml' href='http://www.blogger.com/feeds/7364458407847551072/posts/default/3982372365378327800'/><link rel='alternate' type='text/html' href='http://lpcomp.blogspot.com/2009/03/physical-storage.html' title='Physical storage'/><author><name>David Andersen</name><uri>http://www.blogger.com/profile/03996590425188586871</uri><email>noreply@blogger.com</email><gd:image rel='http://schemas.google.com/g/2005#thumbnail' width='31' height='32' src='http://bp0.blogger.com/_Kj9HEENopvM/SIX7dbPIK4I/AAAAAAAAB_E/39s4kYcj_kc/s1600-R/dave1-sm.jpg'/></author><thr:total>0</thr:total></entry><entry><id>tag:blogger.com,1999:blog-7364458407847551072.post-4631452211677860895</id><published>2009-03-06T11:17:00.001-05:00</published><updated>2009-03-06T11:19:12.519-05:00</updated><title type='text'>Using RFID to prevent ... cactus theft</title><content type='html'>&lt;a href="http://www.treehugger.com/files/2009/03/saguaro-national-park-cactus-cacti-theft-radio-id-chips-rfid.php"&gt;From Treehugger:  Park using RFID chip implants to prevent cactus theft.&lt;/a&gt;  - Desert, low-power chips, lots of sunlight, and burglars.  Who could ask for more?
&lt;img align="right" src="http://www.treehugger.com/Saguaro-cactus-Arizona34534.jpg"&gt;&lt;div class="blogger-post-footer"&gt;&lt;img width='1' height='1' src='https://blogger.googleusercontent.com/tracker/7364458407847551072-4631452211677860895?l=lpcomp.blogspot.com' alt='' /&gt;&lt;/div&gt;</content><link rel='replies' type='application/atom+xml' href='http://lpcomp.blogspot.com/feeds/4631452211677860895/comments/default' title='Post Comments'/><link rel='replies' type='text/html' href='http://lpcomp.blogspot.com/2009/03/using-rfid-to-prevent-cactus-theft.html#comment-form' title='0 Comments'/><link rel='edit' type='application/atom+xml' href='http://www.blogger.com/feeds/7364458407847551072/posts/default/4631452211677860895'/><link rel='self' type='application/atom+xml' href='http://www.blogger.com/feeds/7364458407847551072/posts/default/4631452211677860895'/><link rel='alternate' type='text/html' href='http://lpcomp.blogspot.com/2009/03/using-rfid-to-prevent-cactus-theft.html' title='Using RFID to prevent ... cactus theft'/><author><name>David Andersen</name><uri>http://www.blogger.com/profile/03996590425188586871</uri><email>noreply@blogger.com</email><gd:image rel='http://schemas.google.com/g/2005#thumbnail' width='31' height='32' src='http://bp0.blogger.com/_Kj9HEENopvM/SIX7dbPIK4I/AAAAAAAAB_E/39s4kYcj_kc/s1600-R/dave1-sm.jpg'/></author><thr:total>0</thr:total></entry><entry><id>tag:blogger.com,1999:blog-7364458407847551072.post-4891851444262032343</id><published>2009-02-19T17:47:00.002-05:00</published><updated>2009-02-19T17:49:28.483-05:00</updated><title type='text'>Trust establishment in ISR</title><content type='html'>Arvind Suresh writes:

&lt;blockquote&gt;
&lt;p&gt;
Following the discussions in class yesterday, I thought this would be an appropriate optional reading for those interested in knowing about trust establishment mechanisms in ISR.
&lt;/p&gt;

&lt;p&gt;&lt;a href="http://sparrow.ece.cmu.edu/group/pub/surie-perrig-satya-farber-trustsniffer.pdf"&gt;Rapid Trust Establishment for Pervasive Personal Computing&lt;/a&gt;&lt;br /&gt;
Ajay Surie, Adrian Perrig, Mahadev Satyanarayanan and David J. Farber
&lt;/p&gt;
&lt;/blockquote&gt;&lt;div class="blogger-post-footer"&gt;&lt;img width='1' height='1' src='https://blogger.googleusercontent.com/tracker/7364458407847551072-4891851444262032343?l=lpcomp.blogspot.com' alt='' /&gt;&lt;/div&gt;</content><link rel='replies' type='application/atom+xml' href='http://lpcomp.blogspot.com/feeds/4891851444262032343/comments/default' title='Post Comments'/><link rel='replies' type='text/html' href='http://lpcomp.blogspot.com/2009/02/trust-establishment-in-isr.html#comment-form' title='0 Comments'/><link rel='edit' type='application/atom+xml' href='http://www.blogger.com/feeds/7364458407847551072/posts/default/4891851444262032343'/><link rel='self' type='application/atom+xml' href='http://www.blogger.com/feeds/7364458407847551072/posts/default/4891851444262032343'/><link rel='alternate' type='text/html' href='http://lpcomp.blogspot.com/2009/02/trust-establishment-in-isr.html' title='Trust establishment in ISR'/><author><name>David Andersen</name><uri>http://www.blogger.com/profile/03996590425188586871</uri><email>noreply@blogger.com</email><gd:image rel='http://schemas.google.com/g/2005#thumbnail' width='31' height='32' src='http://bp0.blogger.com/_Kj9HEENopvM/SIX7dbPIK4I/AAAAAAAAB_E/39s4kYcj_kc/s1600-R/dave1-sm.jpg'/></author><thr:total>0</thr:total></entry><entry><id>tag:blogger.com,1999:blog-7364458407847551072.post-4881952135530588370</id><published>2009-02-18T11:30:00.002-05:00</published><updated>2009-02-18T11:34:21.092-05:00</updated><title type='text'>LessWatts.org - Linux power consumption</title><content type='html'>&lt;a href="http://www.cs.cmu.edu/~jfrankli/"&gt;Jason Franklin&lt;/a&gt; points out &lt;a href="http://www.lesswatts.org/"&gt;LessWatts.org&lt;/a&gt;, a very nice repository of power-tuning tricks and tools for Linux.&lt;div class="blogger-post-footer"&gt;&lt;img width='1' height='1' src='https://blogger.googleusercontent.com/tracker/7364458407847551072-4881952135530588370?l=lpcomp.blogspot.com' alt='' /&gt;&lt;/div&gt;</content><link rel='replies' type='application/atom+xml' href='http://lpcomp.blogspot.com/feeds/4881952135530588370/comments/default' title='Post Comments'/><link rel='replies' type='text/html' href='http://lpcomp.blogspot.com/2009/02/lesswattsorg-linux-power-consumption.html#comment-form' title='0 Comments'/><link rel='edit' type='application/atom+xml' href='http://www.blogger.com/feeds/7364458407847551072/posts/default/4881952135530588370'/><link rel='self' type='application/atom+xml' href='http://www.blogger.com/feeds/7364458407847551072/posts/default/4881952135530588370'/><link rel='alternate' type='text/html' href='http://lpcomp.blogspot.com/2009/02/lesswattsorg-linux-power-consumption.html' title='LessWatts.org - Linux power consumption'/><author><name>David Andersen</name><uri>http://www.blogger.com/profile/03996590425188586871</uri><email>noreply@blogger.com</email><gd:image rel='http://schemas.google.com/g/2005#thumbnail' width='31' height='32' src='http://bp0.blogger.com/_Kj9HEENopvM/SIX7dbPIK4I/AAAAAAAAB_E/39s4kYcj_kc/s1600-R/dave1-sm.jpg'/></author><thr:total>0</thr:total></entry><entry><id>tag:blogger.com,1999:blog-7364458407847551072.post-3629817528811089818</id><published>2009-02-11T13:09:00.002-05:00</published><updated>2009-02-11T13:22:49.276-05:00</updated><title type='text'>Obligatory RFID security post:  Cloning passports</title><content type='html'>&lt;p&gt;&lt;a href="http://www.engadget.com/2009/02/02/video-hacker-war-drives-san-francisco-cloning-rfid-passports/"&gt;Another person managed to read passport RFID data remotely&lt;/a&gt;.&lt;/p&gt;

&lt;p&gt;While the topic of the course isn't security, the weakness of many RFID systems &lt;i&gt;is&lt;/i&gt; a direct consequence of extremely-low-power operation:  to avoid the need to have expensive and power-hungry computational capabilities on the RFID chips, many designs use a passive RFID that only sends a single value, instead of being able to participate in a cryptographic protocol.  The results are predictable when they're used in scenarios where reading the chips is possible.  Even some active RFID systems &lt;a href="http://www.usenix.org/events/sec05/tech/bono.html"&gt;take shortcuts that leave them vulnerable to brute-force attacks&lt;/a&gt; (such as using 40-bit keys to reduce computation time and power).&lt;/p&gt;&lt;div class="blogger-post-footer"&gt;&lt;img width='1' height='1' src='https://blogger.googleusercontent.com/tracker/7364458407847551072-3629817528811089818?l=lpcomp.blogspot.com' alt='' /&gt;&lt;/div&gt;</content><link rel='replies' type='application/atom+xml' href='http://lpcomp.blogspot.com/feeds/3629817528811089818/comments/default' title='Post Comments'/><link rel='replies' type='text/html' href='http://lpcomp.blogspot.com/2009/02/obligatory-rfid-security-post-cloning.html#comment-form' title='1 Comments'/><link rel='edit' type='application/atom+xml' href='http://www.blogger.com/feeds/7364458407847551072/posts/default/3629817528811089818'/><link rel='self' type='application/atom+xml' href='http://www.blogger.com/feeds/7364458407847551072/posts/default/3629817528811089818'/><link rel='alternate' type='text/html' href='http://lpcomp.blogspot.com/2009/02/obligatory-rfid-security-post-cloning.html' title='Obligatory RFID security post:  Cloning passports'/><author><name>David Andersen</name><uri>http://www.blogger.com/profile/03996590425188586871</uri><email>noreply@blogger.com</email><gd:image rel='http://schemas.google.com/g/2005#thumbnail' width='31' height='32' src='http://bp0.blogger.com/_Kj9HEENopvM/SIX7dbPIK4I/AAAAAAAAB_E/39s4kYcj_kc/s1600-R/dave1-sm.jpg'/></author><thr:total>1</thr:total></entry><entry><id>tag:blogger.com,1999:blog-7364458407847551072.post-5223152449094546505</id><published>2009-02-10T10:13:00.004-05:00</published><updated>2009-02-10T10:16:36.387-05:00</updated><title type='text'>Google gets into smart power metering</title><content type='html'>&lt;p&gt;In the "using computers to save power elsewhere" arena,
&lt;a href="http://www.google.org/powermeter/"&gt;Google.org has introduced PowerMeter for their employees&lt;/a&gt;.
&lt;/p&gt;

&lt;p&gt;
&lt;a href="http://earth2tech.com/2009/02/09/google-jumps-into-organizing-smart-meter-energy-data/"&gt;Nice writeup from earth2tech.&lt;/a&gt;
&lt;/p&gt;&lt;div class="blogger-post-footer"&gt;&lt;img width='1' height='1' src='https://blogger.googleusercontent.com/tracker/7364458407847551072-5223152449094546505?l=lpcomp.blogspot.com' alt='' /&gt;&lt;/div&gt;</content><link rel='replies' type='application/atom+xml' href='http://lpcomp.blogspot.com/feeds/5223152449094546505/comments/default' title='Post Comments'/><link rel='replies' type='text/html' href='http://lpcomp.blogspot.com/2009/02/google-gets-into-smart-power-metering.html#comment-form' title='0 Comments'/><link rel='edit' type='application/atom+xml' href='http://www.blogger.com/feeds/7364458407847551072/posts/default/5223152449094546505'/><link rel='self' type='application/atom+xml' href='http://www.blogger.com/feeds/7364458407847551072/posts/default/5223152449094546505'/><link rel='alternate' type='text/html' href='http://lpcomp.blogspot.com/2009/02/google-gets-into-smart-power-metering.html' title='Google gets into smart power metering'/><author><name>David Andersen</name><uri>http://www.blogger.com/profile/03996590425188586871</uri><email>noreply@blogger.com</email><gd:image rel='http://schemas.google.com/g/2005#thumbnail' width='31' height='32' src='http://bp0.blogger.com/_Kj9HEENopvM/SIX7dbPIK4I/AAAAAAAAB_E/39s4kYcj_kc/s1600-R/dave1-sm.jpg'/></author><thr:total>0</thr:total></entry><entry><id>tag:blogger.com,1999:blog-7364458407847551072.post-89747313346343458</id><published>2009-01-30T01:15:00.003-05:00</published><updated>2009-01-30T01:18:58.917-05:00</updated><title type='text'>Correction to lecture:  signal strength vs distance</title><content type='html'>First (and certainly not the last) mea culpa update:  I goofed the discussion of the signal strength of an isotropic radiator ("perfect antenna") in class today.  The signal strength degrades with the surface area of the sphere (d^2), not the volume - which is why the path loss exponent in free space is 2.
&lt;br /&gt;

(This correction provided courtesy of the hawk-eyed Charlie Garrod.  Thanks, Charlie!)&lt;div class="blogger-post-footer"&gt;&lt;img width='1' height='1' src='https://blogger.googleusercontent.com/tracker/7364458407847551072-89747313346343458?l=lpcomp.blogspot.com' alt='' /&gt;&lt;/div&gt;</content><link rel='replies' type='application/atom+xml' href='http://lpcomp.blogspot.com/feeds/89747313346343458/comments/default' title='Post Comments'/><link rel='replies' type='text/html' href='http://lpcomp.blogspot.com/2009/01/correction-to-lecture-signal-strength.html#comment-form' title='0 Comments'/><link rel='edit' type='application/atom+xml' href='http://www.blogger.com/feeds/7364458407847551072/posts/default/89747313346343458'/><link rel='self' type='application/atom+xml' href='http://www.blogger.com/feeds/7364458407847551072/posts/default/89747313346343458'/><link rel='alternate' type='text/html' href='http://lpcomp.blogspot.com/2009/01/correction-to-lecture-signal-strength.html' title='Correction to lecture:  signal strength vs distance'/><author><name>David Andersen</name><uri>http://www.blogger.com/profile/03996590425188586871</uri><email>noreply@blogger.com</email><gd:image rel='http://schemas.google.com/g/2005#thumbnail' width='31' height='32' src='http://bp0.blogger.com/_Kj9HEENopvM/SIX7dbPIK4I/AAAAAAAAB_E/39s4kYcj_kc/s1600-R/dave1-sm.jpg'/></author><thr:total>0</thr:total></entry><entry><id>tag:blogger.com,1999:blog-7364458407847551072.post-3148900180747568321</id><published>2009-01-26T21:02:00.002-05:00</published><updated>2009-01-26T21:06:59.543-05:00</updated><title type='text'>Low-Temperature Computing</title><content type='html'>Perhaps antithetical to 15-849, if you happen to have 500 liters of liquid helium to spare, you could &lt;a href="http://www.engadget.com/2009/01/26/amds-phenom-ii-pushed-to-6-5ghz-3dmark-record-demolished/#continued"&gt;try running your CPU at 6.5Ghz at -232 degrees C&lt;/a&gt;.  But in keeping with
what we saw with DVFS, they had to boost the CPU core voltage from about 1.15V to 1.85V.  AMD did not
provide comment on the PUE of a datacenter running on liquid helium.&lt;div class="blogger-post-footer"&gt;&lt;img width='1' height='1' src='https://blogger.googleusercontent.com/tracker/7364458407847551072-3148900180747568321?l=lpcomp.blogspot.com' alt='' /&gt;&lt;/div&gt;</content><link rel='replies' type='application/atom+xml' href='http://lpcomp.blogspot.com/feeds/3148900180747568321/comments/default' title='Post Comments'/><link rel='replies' type='text/html' href='http://lpcomp.blogspot.com/2009/01/low-temperature-computing.html#comment-form' title='0 Comments'/><link rel='edit' type='application/atom+xml' href='http://www.blogger.com/feeds/7364458407847551072/posts/default/3148900180747568321'/><link rel='self' type='application/atom+xml' href='http://www.blogger.com/feeds/7364458407847551072/posts/default/3148900180747568321'/><link rel='alternate' type='text/html' href='http://lpcomp.blogspot.com/2009/01/low-temperature-computing.html' title='Low-Temperature Computing'/><author><name>David Andersen</name><uri>http://www.blogger.com/profile/03996590425188586871</uri><email>noreply@blogger.com</email><gd:image rel='http://schemas.google.com/g/2005#thumbnail' width='31' height='32' src='http://bp0.blogger.com/_Kj9HEENopvM/SIX7dbPIK4I/AAAAAAAAB_E/39s4kYcj_kc/s1600-R/dave1-sm.jpg'/></author><thr:total>0</thr:total></entry><entry><id>tag:blogger.com,1999:blog-7364458407847551072.post-484851901085065690</id><published>2009-01-26T12:37:00.004-05:00</published><updated>2009-01-26T12:56:30.428-05:00</updated><title type='text'>Sun unveils new energy efficient datacenter in Colorado</title><content type='html'>&lt;p&gt;&lt;img align="right" src="https://photos.sun.com/thumbnail/400/400/15888" /&gt;
According to a &lt;a href="http://www.sun.com/aboutsun/pr/2009-01/sunflash.20090126.2.xml"&gt;Sun press release&lt;/a&gt;, they've completed the largest datacenter consolidation project in the company's history -- which they expect to save $1M in electricity costs, and 11,000 metric tons of CO2 per year.  More details in the linked article, but the datacenter uses flywheel-based UPSes, a high-end &lt;a href="http://www.liebert.com/product_pages/SecondaryCategory.aspx?id=33&amp;hz=60"&gt;Liebert XD cooling system&lt;/a&gt; that can dynamically allocate the amount of cooling based on demand.  (Note the little air hoods over each rack.)
&lt;/p&gt;
&lt;p&gt;
The major focus seems to be on space and cooling efficiency, estimating a &lt;a href="http://searchdatacenter.techtarget.com/sDefinition/0,,sid80_gci1307933,00.html"&gt;PUE&lt;/a&gt; of about 1.38.  (For every 1 W of computing power, 1.38 W of power enters the datacenter.) For interesting comparison, Google &lt;a href="http://www.google.com/corporate/green/datacenters/measuring.html"&gt;claims a PUE of 1.21 across six datacenters, with one running at 1.13&lt;/a&gt;.  It's quite likely that some of those datacenters use evaporative cooling in friendly environments, etc.
&lt;/p&gt;&lt;div class="blogger-post-footer"&gt;&lt;img width='1' height='1' src='https://blogger.googleusercontent.com/tracker/7364458407847551072-484851901085065690?l=lpcomp.blogspot.com' alt='' /&gt;&lt;/div&gt;</content><link rel='replies' type='application/atom+xml' href='http://lpcomp.blogspot.com/feeds/484851901085065690/comments/default' title='Post Comments'/><link rel='replies' type='text/html' href='http://lpcomp.blogspot.com/2009/01/sun-unveils-new-energy-efficient.html#comment-form' title='0 Comments'/><link rel='edit' type='application/atom+xml' href='http://www.blogger.com/feeds/7364458407847551072/posts/default/484851901085065690'/><link rel='self' type='application/atom+xml' href='http://www.blogger.com/feeds/7364458407847551072/posts/default/484851901085065690'/><link rel='alternate' type='text/html' href='http://lpcomp.blogspot.com/2009/01/sun-unveils-new-energy-efficient.html' title='Sun unveils new energy efficient datacenter in Colorado'/><author><name>David Andersen</name><uri>http://www.blogger.com/profile/03996590425188586871</uri><email>noreply@blogger.com</email><gd:image rel='http://schemas.google.com/g/2005#thumbnail' width='31' height='32' src='http://bp0.blogger.com/_Kj9HEENopvM/SIX7dbPIK4I/AAAAAAAAB_E/39s4kYcj_kc/s1600-R/dave1-sm.jpg'/></author><thr:total>0</thr:total></entry><entry><id>tag:blogger.com,1999:blog-7364458407847551072.post-8454957130202958532</id><published>2009-01-24T16:52:00.004-05:00</published><updated>2009-01-24T17:01:33.427-05:00</updated><title type='text'>'Net uses more power, more efficiently</title><content type='html'>&lt;a href="http://earth2tech.com/2009/01/23/internet-power-growing-but-becoming-more-energy-efficient/"&gt;Earth2Tech&lt;/a&gt; quotes a report from LBL that explores the growth in Internet energy use (doubling from 2000 to 2006), vs. Internet traffic (increasing by about 22x in the same time period).  A few interesting numbers, like claims that DSL uses 0.17 kWh per gigabyte --- not counting the power draw of your computer, of course.&lt;div class="blogger-post-footer"&gt;&lt;img width='1' height='1' src='https://blogger.googleusercontent.com/tracker/7364458407847551072-8454957130202958532?l=lpcomp.blogspot.com' alt='' /&gt;&lt;/div&gt;</content><link rel='replies' type='application/atom+xml' href='http://lpcomp.blogspot.com/feeds/8454957130202958532/comments/default' title='Post Comments'/><link rel='replies' type='text/html' href='http://lpcomp.blogspot.com/2009/01/net-uses-more-power-more-efficiently.html#comment-form' title='0 Comments'/><link rel='edit' type='application/atom+xml' href='http://www.blogger.com/feeds/7364458407847551072/posts/default/8454957130202958532'/><link rel='self' type='application/atom+xml' href='http://www.blogger.com/feeds/7364458407847551072/posts/default/8454957130202958532'/><link rel='alternate' type='text/html' href='http://lpcomp.blogspot.com/2009/01/net-uses-more-power-more-efficiently.html' title='&apos;Net uses more power, more efficiently'/><author><name>David Andersen</name><uri>http://www.blogger.com/profile/03996590425188586871</uri><email>noreply@blogger.com</email><gd:image rel='http://schemas.google.com/g/2005#thumbnail' width='31' height='32' src='http://bp0.blogger.com/_Kj9HEENopvM/SIX7dbPIK4I/AAAAAAAAB_E/39s4kYcj_kc/s1600-R/dave1-sm.jpg'/></author><thr:total>0</thr:total></entry><entry><id>tag:blogger.com,1999:blog-7364458407847551072.post-6126685662424540941</id><published>2009-01-22T23:36:00.002-05:00</published><updated>2009-01-22T23:41:32.879-05:00</updated><title type='text'>Rackable's "MicroSlice" arrays</title><content type='html'>Companies are starting to introduce &lt;a href="http://searchdatacenter.techtarget.com/news/article/0,289142,sid80_gci1345584,00.html"&gt;products&lt;/a&gt; similar to &lt;a href="http://www.cs.cmu.edu/~fawnproj"&gt;our own research on Fast Arrays of Wimpy Nodes&lt;/a&gt; that exploit the idea of packing more smaller computers into a single enclosure.  The commercial offerings are much less radical than some of the proposed research designs, but should experience some of the benefits anyway.  &lt;a href="http://perspectives.mvdirona.com/"&gt;James Hamilton&lt;/a&gt;, now at Amazon, has a nice post to his (very worthwhile) blog about &lt;a href="http://perspectives.mvdirona.com/2009/01/15/TheCaseForLowCostLowPowerServers.aspx"&gt;low cost, low-power servers&lt;/a&gt;.  The techtarget article attributes to Rackable a statement about getting rid of the need for virtualization, but I suspect that the technologies will remain complimentary.&lt;div class="blogger-post-footer"&gt;&lt;img width='1' height='1' src='https://blogger.googleusercontent.com/tracker/7364458407847551072-6126685662424540941?l=lpcomp.blogspot.com' alt='' /&gt;&lt;/div&gt;</content><link rel='replies' type='application/atom+xml' href='http://lpcomp.blogspot.com/feeds/6126685662424540941/comments/default' title='Post Comments'/><link rel='replies' type='text/html' href='http://lpcomp.blogspot.com/2009/01/rackables-microslice-arrays.html#comment-form' title='0 Comments'/><link rel='edit' type='application/atom+xml' href='http://www.blogger.com/feeds/7364458407847551072/posts/default/6126685662424540941'/><link rel='self' type='application/atom+xml' href='http://www.blogger.com/feeds/7364458407847551072/posts/default/6126685662424540941'/><link rel='alternate' type='text/html' href='http://lpcomp.blogspot.com/2009/01/rackables-microslice-arrays.html' title='Rackable&apos;s &quot;MicroSlice&quot; arrays'/><author><name>David Andersen</name><uri>http://www.blogger.com/profile/03996590425188586871</uri><email>noreply@blogger.com</email><gd:image rel='http://schemas.google.com/g/2005#thumbnail' width='31' height='32' src='http://bp0.blogger.com/_Kj9HEENopvM/SIX7dbPIK4I/AAAAAAAAB_E/39s4kYcj_kc/s1600-R/dave1-sm.jpg'/></author><thr:total>0</thr:total></entry><entry><id>tag:blogger.com,1999:blog-7364458407847551072.post-253911099907257184</id><published>2009-01-20T22:52:00.002-05:00</published><updated>2009-01-20T22:56:50.008-05:00</updated><title type='text'>Symantec:  Data Center Managers Pressured to "Do More with Less"</title><content type='html'>&lt;a href="http://www.symantec.com/about/news/release/article.jsp?prid=20090112_01"&gt;Symantec's State of the Data Center Report&lt;/a&gt; (PR article from Symantec):
&lt;blockquote&gt;
The second annual study found that data center managers are caught between two conflicting goals – more demanding user expectations and higher levels of performance, yet reducing costs remain the primary objective for the &lt;a href="http://www.symantec.com/business/theme.jsp?themeid=datacenter"&gt;data center&lt;/a&gt;. The report also found that data center staffing remains problematic, servers and storage continue to be underutilized and disaster recovery plans are out of date. Finally, the respondents indicated that while they are pursuing green data center initiatives, they are doing so primarily based on cost benefits.
&lt;/blockquote&gt;

And notes that they're attempting to do so using familiar-sounding tools:
&lt;blockquote&gt;
The major server-related initiatives include server consolidation (80 percent) and server virtualization (77 percent).
&lt;/blockquote&gt;

&lt;p&gt;
Or see the full PDF version of the
&lt;a href="http://www.symantec.com/content/en/us/about/media/DataCenter08_Report_Global.pdf"&gt;2008 State of the Data Center report&lt;/a&gt;
(pdf)
&lt;/p&gt;&lt;div class="blogger-post-footer"&gt;&lt;img width='1' height='1' src='https://blogger.googleusercontent.com/tracker/7364458407847551072-253911099907257184?l=lpcomp.blogspot.com' alt='' /&gt;&lt;/div&gt;</content><link rel='replies' type='application/atom+xml' href='http://lpcomp.blogspot.com/feeds/253911099907257184/comments/default' title='Post Comments'/><link rel='replies' type='text/html' href='http://lpcomp.blogspot.com/2009/01/symantec-data-center-managers-pressured.html#comment-form' title='0 Comments'/><link rel='edit' type='application/atom+xml' href='http://www.blogger.com/feeds/7364458407847551072/posts/default/253911099907257184'/><link rel='self' type='application/atom+xml' href='http://www.blogger.com/feeds/7364458407847551072/posts/default/253911099907257184'/><link rel='alternate' type='text/html' href='http://lpcomp.blogspot.com/2009/01/symantec-data-center-managers-pressured.html' title='Symantec:  Data Center Managers Pressured to &quot;Do More with Less&quot;'/><author><name>David Andersen</name><uri>http://www.blogger.com/profile/03996590425188586871</uri><email>noreply@blogger.com</email><gd:image rel='http://schemas.google.com/g/2005#thumbnail' width='31' height='32' src='http://bp0.blogger.com/_Kj9HEENopvM/SIX7dbPIK4I/AAAAAAAAB_E/39s4kYcj_kc/s1600-R/dave1-sm.jpg'/></author><thr:total>0</thr:total></entry><entry><id>tag:blogger.com,1999:blog-7364458407847551072.post-4561484695376745617</id><published>2009-01-20T18:02:00.005-05:00</published><updated>2009-01-20T22:58:38.252-05:00</updated><title type='text'>Intel articles about low-power CPU states</title><content type='html'>Anshul writes with two relevant links for low power states in CPUs:

&lt;p&gt;
&lt;a href="http://download.intel.com/technology/itj/2006/volume10issue02/vol10_art03.pdf"&gt;2006 article about P and C states in Intel processors&lt;/a&gt;.
&lt;/p&gt;

&lt;p&gt;The second article
&lt;a href="http://download.intel.com/technology/eep/ecorackwp.pdf"&gt;talks about the C1E (extended HLT) state&lt;/a&gt;,
(Page 6)
which is like normal C1 but also turns voltage down to the lowest level.  This type of sleep state may address some of the issues of using idle states in the first two papers we read this semester.
&lt;/p&gt;&lt;div class="blogger-post-footer"&gt;&lt;img width='1' height='1' src='https://blogger.googleusercontent.com/tracker/7364458407847551072-4561484695376745617?l=lpcomp.blogspot.com' alt='' /&gt;&lt;/div&gt;</content><link rel='replies' type='application/atom+xml' href='http://lpcomp.blogspot.com/feeds/4561484695376745617/comments/default' title='Post Comments'/><link rel='replies' type='text/html' href='http://lpcomp.blogspot.com/2009/01/anshul-writes-with-two-relevant-links.html#comment-form' title='2 Comments'/><link rel='edit' type='application/atom+xml' href='http://www.blogger.com/feeds/7364458407847551072/posts/default/4561484695376745617'/><link rel='self' type='application/atom+xml' href='http://www.blogger.com/feeds/7364458407847551072/posts/default/4561484695376745617'/><link rel='alternate' type='text/html' href='http://lpcomp.blogspot.com/2009/01/anshul-writes-with-two-relevant-links.html' title='Intel articles about low-power CPU states'/><author><name>David Andersen</name><uri>http://www.blogger.com/profile/03996590425188586871</uri><email>noreply@blogger.com</email><gd:image rel='http://schemas.google.com/g/2005#thumbnail' width='31' height='32' src='http://bp0.blogger.com/_Kj9HEENopvM/SIX7dbPIK4I/AAAAAAAAB_E/39s4kYcj_kc/s1600-R/dave1-sm.jpg'/></author><thr:total>2</thr:total></entry><entry><id>tag:blogger.com,1999:blog-7364458407847551072.post-3625770698962460061</id><published>2009-01-20T01:18:00.000-05:00</published><updated>2009-01-20T01:20:19.324-05:00</updated><title type='text'>Cisco entering virtual server market</title><content type='html'>Relevant to today's readings in virtualization for cluster energy savings:  &lt;a href="http://www.nytimes.com/2009/01/20/technology/companies/20cisco.html?scp=1&amp;amp;sq=cisco%20server&amp;amp;st=cse"&gt;Cisco Plans Big Push into Server Market&lt;/a&gt;.
&lt;blockquote&gt;
“Our vision is, how do we virtualize the entire data center?” Ms. Warrior said. “It is not about a single product. We will have a series of products that enable us to make that transition.”
&lt;/blockquote&gt;&lt;div class="blogger-post-footer"&gt;&lt;img width='1' height='1' src='https://blogger.googleusercontent.com/tracker/7364458407847551072-3625770698962460061?l=lpcomp.blogspot.com' alt='' /&gt;&lt;/div&gt;</content><link rel='replies' type='application/atom+xml' href='http://lpcomp.blogspot.com/feeds/3625770698962460061/comments/default' title='Post Comments'/><link rel='replies' type='text/html' href='http://lpcomp.blogspot.com/2009/01/cisco-entering-virtual-server-market.html#comment-form' title='0 Comments'/><link rel='edit' type='application/atom+xml' href='http://www.blogger.com/feeds/7364458407847551072/posts/default/3625770698962460061'/><link rel='self' type='application/atom+xml' href='http://www.blogger.com/feeds/7364458407847551072/posts/default/3625770698962460061'/><link rel='alternate' type='text/html' href='http://lpcomp.blogspot.com/2009/01/cisco-entering-virtual-server-market.html' title='Cisco entering virtual server market'/><author><name>David Andersen</name><uri>http://www.blogger.com/profile/03996590425188586871</uri><email>noreply@blogger.com</email><gd:image rel='http://schemas.google.com/g/2005#thumbnail' width='31' height='32' src='http://bp0.blogger.com/_Kj9HEENopvM/SIX7dbPIK4I/AAAAAAAAB_E/39s4kYcj_kc/s1600-R/dave1-sm.jpg'/></author><thr:total>0</thr:total></entry></feed>
