Official: AMD leads by 50%
AMD officially confirmed that K10 will be 50% faster than Intel's fastest quad core.
That's a staggering and crushing performance advantage.
Both AMD and Intel CPUs are roughly at the same clockspeed at about 3GHZ, as both are limited by the same physics and manufacturing equipment they purchased from tool companies, such as Applied Materials.
Below the top speed, there are about 10 speed grades. The fastest is about 50% faster than the slowest. Thus, each grade is about 5% faster than the one below.
On desktop, Intel now has about 10-12% IPC advantage with Core 2. This puts AMD's X2 6000+ at roughly the same speed as Intel's E6700. This situation has caused tremendous problems for AMD.
Now, with 50% performance advantage, the slowest K10 CPU will be faster than the fastest Intel CPU.
It's reported that Intel execs have exercised their options. Congrats to Patty. At least, he can make some money before becoming jobless next year.
posted by Sharikou, Ph. D. @ 4/23/2007 12:56:00 PM
16 Comments:
Sharikou
I have to ask did you even read the press release?
"The new Barcelona projections are based on the latest SPECcpu2006 benchmarks and show that AMD expects to have up to a 50 percent advantage in floating point performance and 20 percent in integer performance over the competition’s highest-performing quad-core processor at the same frequency".
Considering they will not be at the same frequency, the 20% Integer advantage should/could be fairly easy for Intel to overcome with higher clock speeds.
Floating point is still limited by the FSB, while HTT continues to be AMD's advantage.
It should be interesting to see what happens with the 1600 FSB Intel is supposedly going to release and how the added bandwidth effects memory intensive testing.
Barcelona has vastly superior system bandwidth compared to Clowertown. It is a miracle how it only achieves 50% better performance on bandwidth limited tests like specfp, same benchmark that this 50% number is based on
Almost forgot,
"K10 core will have 40% IPC lead over Core 2 on integer (per core).
K10 will have 200% IPC lead over Core 2 on floating point (per core)."
Now official numbers based on one single synthetic benchmark show only 20 and 50%. Kind of odd, isn't it?
Official links? None? Then as usual you spout official bullshit.
There's a reason why AMD put NDA's on the press at the so called launch event today...if you stop and think about it, you'd have half a clue...
In related news, I am UP TO 50% better golfer then Tiger Woods.
"The new Barcelona projections are based on the latest SPECcpu2006 benchmarks and show that AMD expects to have up to a 50 percent advantage in floating point performance and 20 percent in integer performance over the competition’s highest-performing quad-core processor at the same frequency."
So AMD 'expects' Barcelona to be better in a single benchmark. I'm mot impressed. Where are the benchmarks, where are details on real application performance, where is the processor ???
AMD will be between 20 and 50% above intel's CURRENT processors. Penryn will be what, about 40% above their current processors? Looks like it will be a tight race. The only thing is, is that Intel has SHOWN benchmarks (although perhaps slightly biased) while AMD has only SAID things...has anyone seen any benchmarks that show 20-50% increase?
SHOCKER: SHARIKOU ADMITS AMD BEING IS FRAGGED TODAY BY INTEL!
"This puts AMD's X2 6000+ at roughly the same speed as Intel's E6700. This situation has caused tremendous problems for AMD."
this is so unprecedented that i'm not even going to point out to him that INTEL is STILL causing AMD tremendous problems. Barcelona ships later Q3. Paper launched benchmarks don't generate revenues.
@Scott
Where is your brain at, are you a computer bot, or you purpousely dumb for NO reason . Anandtech has done some early benchmarks and they proved that there is barely %10 advantage and it was because of 6mbs of cache. Pathetic!!!
I still can't believe some people have low IQ like intelers. Patty said,"Penrym will have 40% advantage over the current line of Core 2 Duos". Do you still believe in that? I don't see the 40% according to Anandtech. Patty is MAD and RED.
Intel knows that Barcelona will outperform the current line of Core 2 Duos.
http://www.anandtech.com/cpuchipsets/intel/showdoc.aspx?i=2972&p=2
I have heard that once HT3 comes online it will add another 50% performance to the already purposed 50% edge. Talk about superior architecture if true.
So, tommy, AMD isn't adding sort of cache to Barcelona is it? After all, apparently, that's the easy way out.
The only performance claims that we've seen from AMD are from Randy Allen at AMD, who claimed Barcelona would be faster by 40% in a "wide variety of workloads" and now this press release.
"Penryn vs. Conroe improvements to be in the 5 - 10% range" according to anandtech.
"Barcelona vs Conroe improvements to be in the 40-50% range" according to amd officially report.
So, it can be concluded that..
"Barcelona to be faster thn penryn about 35-40% range"
and also..
Barcelona (8 Cores) to be faster thn Penryn (4 cores) about 70-80%.. Ekekekekekekeke.. Go amd go... go.. go..
giant - That wasn't my main concern. I was hoping to point out that Intel simply lied to people or to its fans. I mean crap... Why would a huge company like INTEL claim something that is false??? Someone may think for marketing purpouses or something else. I don't know... perhaps someone can give me a brief explaination of that.
Looking at the 10% Anand said for Penryn that was a reduced clock for clock comparison as the Penryn numbers were at 3.33 not 2.93GHz. The 40% was when SSE4 could be utilized.
In summary Penryn will likely overclock better and have a slightly better IPC, whether that is enough to beat off K10 is unknown, maybe yes, maybe no.
I would imagine that Intel are hoping to be able to release a QX 333x11 part or perhaps 400x9 and hoping that AMD cannot get anywhere near 3GHz for general release. If K10 can go out the door at 3GHz then it looks pretty rosy for the green camp.
r
"I have heard that once HT3 comes online it will add another 50% performance to the already purposed 50% edge"
It will only help on 2P and upwards in situations where you need lots of bandwidth between CPUs. It will not help 1P PC's and servers.
pezal
"according to amd officially report."
If Intel reported Core2 was 40/40 architecture nobody believed it. Now that AMD throws out some numbers without any proof it must be taken as gospel.
tommy
"was hoping to point out that Intel simply lied to people or to its fans"
Just calculate the average improvement for those benchmarks. DivX encoding gives it a huge boost and average comes out just around what Intel said. It isn't any more lying than AMD claiming 40% more benchmarks in variety of workloads and then saying only 20% improvement in integer throughput.
amw
"If K10 can go out the door at 3GHz then it looks pretty rosy for the green camp."
AMD won't have >2.5GHz quads this year. Intel will have >3.2GHz quads this year.
You know, I love this site; it really is entertaining during a long workday. I know most of the site is bias rhetoric, but occasionally you will find a glimmer of truth or at least some interesting links to view. I must say that this claimed performance increase of 50ish percent, if true; is completely amazing. I do have my doubts though, granted AMD has not really led us astray before, but I am still skeptical.
I am a huge AMD fan, but I am seriously considering buying an Intel laptop at this point. I am going to hold off to see what the AMD camp has, but given the fact their laptop processor is probably half a year away I doubt I will wait for its release.
Post a Comment
<< Home