<?xml version="1.0" encoding="utf-8"?>
<rss version="2.0" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:atom="http://www.w3.org/2005/Atom">
	<channel>
		<title>Asus leaks more Bulldozer specs</title>
		<description>Discuss Asus leaks more Bulldozer specs</description>
		<link>http://fudzilla.com/home/item/22808-asus-leaks-more-bulldozer-specs</link>
		<lastBuildDate>Fri, 24 May 2013 03:33:32 +0100</lastBuildDate>
		<generator>JComments</generator>
		<atom:link href="http://www.fudzilla.com/component/jcomments/feed/com_k2/22808/10" rel="self" type="application/rss+xml" />
		<item>
			<title>mikato says:</title>
			<link>http://fudzilla.com/home/item/22808-asus-leaks-more-bulldozer-specs#comment-23456</link>
			<description><![CDATA[The definition of core is fuzzying since AMD chose to call 1 BD module as 2 cores instead of going w/ 1 BD module=1 core. What it boils down to is changing the mix of the execution types (int/fpu). AMD is banking on more use of the int units of CPUs (long term likely banking on fully integrated GPU taking most duties of fp so maybe we'll see no fpus in the CPU at some point). With this idea they'll have less idle fpus in the CPU, save on die space, therefore manuf. cost & end price. It also saves on CPU heat & makes CPU more efficient since more is being used. An all around win for an arch change if that's how our CPUs are used. Proof is in the pudding so we'll see what comes out. Even if it doesn't catch AMD up in benchmarks, it could still be a shift that gives an edge long term.]]></description>
			<dc:creator>mikato</dc:creator>
			<pubDate>Thu, 26 May 2011 16:23:04 +0100</pubDate>
			<guid>http://fudzilla.com/home/item/22808-asus-leaks-more-bulldozer-specs#comment-23456</guid>
		</item>
		<item>
			<title>mikato says:</title>
			<link>http://fudzilla.com/home/item/22808-asus-leaks-more-bulldozer-specs#comment-23455</link>
			<description><![CDATA[Sure :) Each Bulldozer module can handle two threads - it has all the infrastructure (decode and instruction issue as you say) to send them through completely in parallel. Each module has two integer units and one fp unit so it will be sending the threads to use two out of those three under load. This is in contrast to Intel's HyperThreading which has infrastructure to handle two threads with one core but only one can actually be executed at a time (meaning one out of two is used of the integer and fpu).]]></description>
			<dc:creator>mikato</dc:creator>
			<pubDate>Thu, 26 May 2011 16:11:28 +0100</pubDate>
			<guid>http://fudzilla.com/home/item/22808-asus-leaks-more-bulldozer-specs#comment-23455</guid>
		</item>
		<item>
			<title>Naterm says:</title>
			<link>http://fudzilla.com/home/item/22808-asus-leaks-more-bulldozer-specs#comment-23417</link>
			<description><![CDATA[No, you cannot 'just use bigger heatsinks'. That's not even really possible with desktop computers, it's definitely not possible with servers. You can only jam so much heatsink into a rack chassis or blade. Intel and AMD should use the same tests, but AMD isn't going to do that. They'd look even worse than they do now. TDP is a maximum figure, it will rarely be met, especially in consumer systems. Now on a heavily consolidated server or an HPC node, you'll likely get close. That said, there is little doubt who is more power efficient. It's not AMD and hasn't been since 2006.]]></description>
			<dc:creator>Naterm</dc:creator>
			<pubDate>Thu, 26 May 2011 03:21:27 +0100</pubDate>
			<guid>http://fudzilla.com/home/item/22808-asus-leaks-more-bulldozer-specs#comment-23417</guid>
		</item>
		<item>
			<title>hoohoo says:</title>
			<link>http://fudzilla.com/home/item/22808-asus-leaks-more-bulldozer-specs#comment-23326</link>
			<description><![CDATA[Be that as it may: what exactly does TDP mean in quantitative terms? It cannot be how many watts before the case melts - we can always use bigger heat sinks. It cannot be maximum heat generated by the device because that's what TPD is. TDP is an ambiguous statistic. If TDP is merely an envelope within which the chip will operate, that's OK, but INTC & AMD should measure it the same way. That TDP means nothing is ultimately attested to by that we all end up talking about what the power consumption is measured to be in real tests!]]></description>
			<dc:creator>hoohoo</dc:creator>
			<pubDate>Tue, 24 May 2011 21:57:28 +0100</pubDate>
			<guid>http://fudzilla.com/home/item/22808-asus-leaks-more-bulldozer-specs#comment-23326</guid>
		</item>
		<item>
			<title>hoohoo says:</title>
			<link>http://fudzilla.com/home/item/22808-asus-leaks-more-bulldozer-specs#comment-23325</link>
			<description><![CDATA[Having been into computers since circa 1988, and electronics since about 1975, I can only say I never ever saw 'TDP' quoted until Athlons and P4 hit the 100+ watt mark and the two companies started marketing on power consumption as well as MHz. 'Total Power Dissipation (max)' was typical language on electronic part spec sheets.]]></description>
			<dc:creator>hoohoo</dc:creator>
			<pubDate>Tue, 24 May 2011 21:51:55 +0100</pubDate>
			<guid>http://fudzilla.com/home/item/22808-asus-leaks-more-bulldozer-specs#comment-23325</guid>
		</item>
		<item>
			<title>Naterm says:</title>
			<link>http://fudzilla.com/home/item/22808-asus-leaks-more-bulldozer-specs#comment-23321</link>
			<description><![CDATA[I can't say about the consumer processors, but I can comment on the server chips. Intel's TDP specs for TDP are more accurate and have been for several years. Intel is also much more power efficient than AMD if you look at actual benchmarks.]]></description>
			<dc:creator>Naterm</dc:creator>
			<pubDate>Tue, 24 May 2011 21:09:40 +0100</pubDate>
			<guid>http://fudzilla.com/home/item/22808-asus-leaks-more-bulldozer-specs#comment-23321</guid>
		</item>
		<item>
			<title>Naterm says:</title>
			<link>http://fudzilla.com/home/item/22808-asus-leaks-more-bulldozer-specs#comment-23320</link>
			<description><![CDATA[No, it's always been 'Thermal Design Power'.]]></description>
			<dc:creator>Naterm</dc:creator>
			<pubDate>Tue, 24 May 2011 21:07:07 +0100</pubDate>
			<guid>http://fudzilla.com/home/item/22808-asus-leaks-more-bulldozer-specs#comment-23320</guid>
		</item>
		<item>
			<title>Bl0bb3r says:</title>
			<link>http://fudzilla.com/home/item/22808-asus-leaks-more-bulldozer-specs#comment-23268</link>
			<description><![CDATA[That's why power usage tests don't take TDP into consideration... I've seen Athlon II x4 rated 95W TDP but wusing a mere 55W.]]></description>
			<dc:creator>Bl0bb3r</dc:creator>
			<pubDate>Tue, 24 May 2011 11:25:33 +0100</pubDate>
			<guid>http://fudzilla.com/home/item/22808-asus-leaks-more-bulldozer-specs#comment-23268</guid>
		</item>
		<item>
			<title>hoohoo says:</title>
			<link>http://fudzilla.com/home/item/22808-asus-leaks-more-bulldozer-specs#comment-23251</link>
			<description><![CDATA[Indeed. Back in the days before marketers wrote technical specs we had Total Power Dissipation == how many watts does the d*mn thing dissipate maximum. Now we have TDP == Total Dissipation Under Whatever Scenario Some Kook In Marketing Thinks Will Con The Punters.]]></description>
			<dc:creator>hoohoo</dc:creator>
			<pubDate>Tue, 24 May 2011 08:21:49 +0100</pubDate>
			<guid>http://fudzilla.com/home/item/22808-asus-leaks-more-bulldozer-specs#comment-23251</guid>
		</item>
		<item>
			<title>hoohoo says:</title>
			<link>http://fudzilla.com/home/item/22808-asus-leaks-more-bulldozer-specs#comment-23250</link>
			<description><![CDATA[But... what is not clear to me is whether the decode and instruction issue in a BD module is capable of driving the 2 ALUs and the FPU with completely arbitrary instruction streams. Perhaps someone else can inform/correct/flame me? :-)]]></description>
			<dc:creator>hoohoo</dc:creator>
			<pubDate>Tue, 24 May 2011 08:17:39 +0100</pubDate>
			<guid>http://fudzilla.com/home/item/22808-asus-leaks-more-bulldozer-specs#comment-23250</guid>
		</item>
	</channel>
</rss>
