<?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/"
	xmlns:go='http://ns.gigaom.com/'
xmlns:georss="http://www.georss.org/georss" xmlns:geo="http://www.w3.org/2003/01/geo/wgs84_pos#" xmlns:media="http://search.yahoo.com/mrss/"
		>
<channel>
	<title>Comments on: Cisco&#039;s New Router Shows Need for New Processors</title>
	<atom:link href="http://gigaom.com/2008/11/11/ciscos-new-router-shows-need-for-new-processors/feed/" rel="self" type="application/rss+xml" />
	<link>http://gigaom.com/2008/11/11/ciscos-new-router-shows-need-for-new-processors/</link>
	<description></description>
	<lastBuildDate>Fri, 10 Feb 2012 21:40:26 +0000</lastBuildDate>
	<sy:updatePeriod>hourly</sy:updatePeriod>
	<sy:updateFrequency>1</sy:updateFrequency>
	<generator>http://wordpress.com/</generator>
	<item>
		<title>By: till</title>
		<link>http://gigaom.com/2008/11/11/ciscos-new-router-shows-need-for-new-processors/#comment-152131</link>
		<dc:creator><![CDATA[till]]></dc:creator>
		<pubDate>Thu, 13 Nov 2008 20:07:11 +0000</pubDate>
		<guid isPermaLink="false">http://gigaom.com/?p=28805#comment-152131</guid>
		<description><![CDATA[Maybe the processors are really not the issue here, maybe it&#039;s the weird OS cisco runs on top of it?

There are switches in telco that guarantee 99.999999999% availability (fraction of a second of downtime per year), and those switches have been in service for 10 years. Point taken they are used to handle mostly phone-related, but who says the technology that has been around all those years (or maybe more), can&#039;t handle Internet traffic?

Imagine what today&#039;s hardware could do if manufactures used it to the fullest? If you don&#039;t believe me, think about all those dumb devices sold by d-link, netgear and linksys - put a linux on it and you have a very capable device. Point taken that kind of hardware does not handle TB/s but I&#039;m sure you can follow my analogy.

By the way, the switches (with the nine &quot;nines&quot;) are by Ericsson and the runtime on them is Erlang. Not as wide spread and popular, but for example projects such as eJabberd, CouchDB, etc. make use of it and demonstrate what&#039;s hidden in the hardware.

A lot of questions, none answered by a cisco press release. :-)]]></description>
		<content:encoded><![CDATA[<p>Maybe the processors are really not the issue here, maybe it&#8217;s the weird OS cisco runs on top of it?</p>
<p>There are switches in telco that guarantee 99.999999999% availability (fraction of a second of downtime per year), and those switches have been in service for 10 years. Point taken they are used to handle mostly phone-related, but who says the technology that has been around all those years (or maybe more), can&#8217;t handle Internet traffic?</p>
<p>Imagine what today&#8217;s hardware could do if manufactures used it to the fullest? If you don&#8217;t believe me, think about all those dumb devices sold by d-link, netgear and linksys &#8211; put a linux on it and you have a very capable device. Point taken that kind of hardware does not handle TB/s but I&#8217;m sure you can follow my analogy.</p>
<p>By the way, the switches (with the nine &#8220;nines&#8221;) are by Ericsson and the runtime on them is Erlang. Not as wide spread and popular, but for example projects such as eJabberd, CouchDB, etc. make use of it and demonstrate what&#8217;s hidden in the hardware.</p>
<p>A lot of questions, none answered by a cisco press release. :-)</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Stacey Higginbotham</title>
		<link>http://gigaom.com/2008/11/11/ciscos-new-router-shows-need-for-new-processors/#comment-152130</link>
		<dc:creator><![CDATA[Stacey Higginbotham]]></dc:creator>
		<pubDate>Wed, 12 Nov 2008 18:45:45 +0000</pubDate>
		<guid isPermaLink="false">http://gigaom.com/?p=28805#comment-152130</guid>
		<description><![CDATA[Jakob, in my effort to simplify and general server-centric worldview I should have specified that I&#039;m not talking about a variety DSPs but actually identical processor cores  as they do on servers. As for the GPU point, that&#039;s exactly what I&#039;m trying to highlight. For embedded apps we need to think about performance and power.]]></description>
		<content:encoded><![CDATA[<p>Jakob, in my effort to simplify and general server-centric worldview I should have specified that I&#8217;m not talking about a variety DSPs but actually identical processor cores  as they do on servers. As for the GPU point, that&#8217;s exactly what I&#8217;m trying to highlight. For embedded apps we need to think about performance and power.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Jakob Engblom</title>
		<link>http://gigaom.com/2008/11/11/ciscos-new-router-shows-need-for-new-processors/#comment-152129</link>
		<dc:creator><![CDATA[Jakob Engblom]]></dc:creator>
		<pubDate>Wed, 12 Nov 2008 13:41:50 +0000</pubDate>
		<guid isPermaLink="false">http://gigaom.com/?p=28805#comment-152129</guid>
		<description><![CDATA[&gt; Instead of making chips for such devices more powerful (and more power-hungry), engineers are following in the footsteps of the server world and adding more cores.

Well, it is the other way around. The embedded multicore chips predates the general-purpose ones by about five years.  The first shipping embedded multicore came in 1994 (TI video chip with 4 DSP cores), the first general one was really launched in 2001 with Power4.  And if you count the Motorola QUICC design, it goes back to 1991.

&gt; Multicore chips are gaining in use in the embedded world for networking gear, set-top boxes and other applications.

You forget mobile phones, which have been multicore for many years now. I doubt a single phone shipped in the last three years have contained single-processor chips (apart from some quite specific ASICs for peripheral functionality).

&gt; Tensilica’s DPU cores and Freescale’s chips offer a way to process that information using less power than a general purpose CPU or even a graphics processor that might also be used for the task.

That shows some ignorance of embedded multicore... a GPU is a ridiculously inefficient beast built for maximum performance rather than good performance/power ratio. Tensilicas cores are maybe not as fast as the GPU core, but they sure run faster per watt.

/jakob]]></description>
		<content:encoded><![CDATA[<p>&gt; Instead of making chips for such devices more powerful (and more power-hungry), engineers are following in the footsteps of the server world and adding more cores.</p>
<p>Well, it is the other way around. The embedded multicore chips predates the general-purpose ones by about five years.  The first shipping embedded multicore came in 1994 (TI video chip with 4 DSP cores), the first general one was really launched in 2001 with Power4.  And if you count the Motorola QUICC design, it goes back to 1991.</p>
<p>&gt; Multicore chips are gaining in use in the embedded world for networking gear, set-top boxes and other applications.</p>
<p>You forget mobile phones, which have been multicore for many years now. I doubt a single phone shipped in the last three years have contained single-processor chips (apart from some quite specific ASICs for peripheral functionality).</p>
<p>&gt; Tensilica’s DPU cores and Freescale’s chips offer a way to process that information using less power than a general purpose CPU or even a graphics processor that might also be used for the task.</p>
<p>That shows some ignorance of embedded multicore&#8230; a GPU is a ridiculously inefficient beast built for maximum performance rather than good performance/power ratio. Tensilicas cores are maybe not as fast as the GPU core, but they sure run faster per watt.</p>
<p>/jakob</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Olaf Baumert</title>
		<link>http://gigaom.com/2008/11/11/ciscos-new-router-shows-need-for-new-processors/#comment-152128</link>
		<dc:creator><![CDATA[Olaf Baumert]]></dc:creator>
		<pubDate>Wed, 12 Nov 2008 07:28:53 +0000</pubDate>
		<guid isPermaLink="false">http://gigaom.com/?p=28805#comment-152128</guid>
		<description><![CDATA[Excuse my spelling, just got up *yawn*]]></description>
		<content:encoded><![CDATA[<p>Excuse my spelling, just got up *yawn*</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Olaf Baumert</title>
		<link>http://gigaom.com/2008/11/11/ciscos-new-router-shows-need-for-new-processors/#comment-152127</link>
		<dc:creator><![CDATA[Olaf Baumert]]></dc:creator>
		<pubDate>Wed, 12 Nov 2008 07:27:37 +0000</pubDate>
		<guid isPermaLink="false">http://gigaom.com/?p=28805#comment-152127</guid>
		<description><![CDATA[Maybe you shoould compute the 6.4 Terrabytes (per second! don&#039;t forget this) into something real ;).
Like gbit/s. Than throw away half of this. As router vendors tend to double there numbers (&quot;But it&#039;s sending! And receiving!&quot;)

&quot;old processors just can’t cut it anymore&quot; says who? :-)]]></description>
		<content:encoded><![CDATA[<p>Maybe you shoould compute the 6.4 Terrabytes (per second! don&#8217;t forget this) into something real ;).<br />
Like gbit/s. Than throw away half of this. As router vendors tend to double there numbers (&#8220;But it&#8217;s sending! And receiving!&#8221;)</p>
<p>&#8220;old processors just can’t cut it anymore&#8221; says who? :-)</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Om Malik</title>
		<link>http://gigaom.com/2008/11/11/ciscos-new-router-shows-need-for-new-processors/#comment-152126</link>
		<dc:creator><![CDATA[Om Malik]]></dc:creator>
		<pubDate>Wed, 12 Nov 2008 00:07:48 +0000</pubDate>
		<guid isPermaLink="false">http://gigaom.com/?p=28805#comment-152126</guid>
		<description><![CDATA[@foobar .. how about you being a man enough to leave your name instead of hiding behind foobar !]]></description>
		<content:encoded><![CDATA[<p>@foobar .. how about you being a man enough to leave your name instead of hiding behind foobar !</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: foobar</title>
		<link>http://gigaom.com/2008/11/11/ciscos-new-router-shows-need-for-new-processors/#comment-152125</link>
		<dc:creator><![CDATA[foobar]]></dc:creator>
		<pubDate>Tue, 11 Nov 2008 23:17:14 +0000</pubDate>
		<guid isPermaLink="false">http://gigaom.com/?p=28805#comment-152125</guid>
		<description><![CDATA[Don&#039;t you guys want to stick to blogging about web2.0 things such as blogs and startups?]]></description>
		<content:encoded><![CDATA[<p>Don&#8217;t you guys want to stick to blogging about web2.0 things such as blogs and startups?</p>
]]></content:encoded>
	</item>
</channel>
</rss>

