Sign in with
Sign up | Sign in
Your question

Intel 8 core faster than amd 16 core?

Last response: in CPUs
Share
May 23, 2007 4:27:05 PM

Lets see Sharikou try and spin this into good news! :lol:  8 Core Clovertown system frags 16 core Barcelona system by 30%! :trophy: :trophy: :trophy:
May 23, 2007 4:33:47 PM

I'd change the title of your post to "Intel Xeon 8 cores beat Barcy 16 cores"

It's just not clear what your post is about.

---
something has got to be wrong w/ the test... something....
Related resources
May 23, 2007 4:37:47 PM

The question is about a 3ghz system against a 2 ghz? system. Does anyone know what speed the AMD system is running at when making this comparison?
May 23, 2007 4:43:23 PM

Randy Allen did not specify the clockspeed. He only said that it wasn't the fastest quad core they will have on the market when they release Barcelona.
May 23, 2007 4:45:00 PM

Quote:
The question is about a 3ghz system against a 2 ghz? system. Does anyone know what speed the AMD system is running at when making this comparison?


People are making wild guesses on other forums. I'm reading some people think it's as low as 1ghz or that it's the shared cache that's killing it.
a b à CPUs
May 23, 2007 4:48:10 PM

This is funny..lol
May 23, 2007 4:55:30 PM

Quote:
something has got to be wrong w/ the test... something....


Or AMD benched with really low clocks, either as a result of problems or trying to hide.

http://www.lostcircuits.com/cpu/amd_quadfx/11.shtml

(1.6/3)*2139=a score of 1140.8 for 4 1.6GHz K8 cores. Times two for 8 cores yields ~2281.
May 23, 2007 4:56:05 PM

Hardware vs vaporware and the vaporware got smoked incredible. 8O

AMD is in a serious world of hurt if Barcy can't beat the current Xeon cores and Penryn is going to be another 40% faster. :cry:  :cry: 
May 23, 2007 4:59:59 PM

Don't make too much out of that 40% faster score. I think a lot of that had to do with SSE4 instruction set that Penryn can allow programs utilize. It's got a funtion to allow better rendering of high definition video. I believe they probably used the 3.33 ghz model Penryn as well, giving it a slight boost over the QX6800's 2.93 ghz.
May 23, 2007 5:03:18 PM

Ok... I'm hearing that it was just a driver issue...


:lol:  ( j/k couldn't resist)
May 23, 2007 5:23:58 PM

the video encoding demo in the video is really cool, penryn is gonna be awesome.
May 23, 2007 5:25:43 PM

Are you familiar with what software was encoding?
a b à CPUs
May 23, 2007 5:42:55 PM

Didn't we see things like this before where there was a bunch of tit-for-tat benchmarks and both Intel and AMDbested each other and then claimed that the other was not running an "optimized" system with outdated drivers? Yeah, I thought we had...

Well, here we go again! :roll:
May 23, 2007 6:06:47 PM

Quote:
Didn't we see things like this before where there was a bunch of tit-for-tat benchmarks and both Intel and AMDbested each other and then claimed that the other was not running an "optimized" system with outdated drivers? Yeah, I thought we had...

Well, here we go again! :roll:


Yeah, cause that's almost clinically impossible. Barcelona CAN'T be slower than Opteron.
Just wierd.
May 23, 2007 6:15:03 PM

Woohoo!! Come on P35 boards!
May 23, 2007 6:17:40 PM

I have no idea how barcelona will perform, but here are 3 questiions that maybe Baron or somebody can explain to me.

1) - AMDs sales have declined prettty sharply, if Barcelona rocked, would it not make sense to release those numbers, and at least hold some market share by showing that 3 months out they have a better "mousetrap" to sell..? - "Buy a x2 3600+ and in 3 months drop in a Barcelona that rocks, here is the proof" strikes me as a better sales pitch than.. "It rocks, really it does, take my word for it... would you like a 4600+ while you are waiting...?"

2) - AMD has recently been required to go back to the capital markets to raise 1+ $Billion. If they do have a CPU that kicks butt, would they not have been able to secure better rates/terms from the market if they showed a return to performance leadership as opposed to raising cash from a position of percieved technological inferiority?

3) - Generically - why play things so close to the vest? - Even if Barcelona truly does rock, what is gained by hiding it?
May 23, 2007 6:17:41 PM

Quote:
Didn't we see things like this before where there was a bunch of tit-for-tat benchmarks and both Intel and AMDbested each other and then claimed that the other was not running an "optimized" system with outdated drivers? Yeah, I thought we had...

Well, here we go again! :roll:


Yeah, cause that's almost clinically impossible. Barcelona CAN'T be slower than Opteron.
Just wierd.

Sure it can, if they can't yield **** for clockspeed. They seem to be having quite the problem getting Brisbane clockspeeds up to the Windsor ones. It's coming up on 6 months since the Brisbane launch:

http://dailytech.com/article.aspx?newsid=5222

and where are the 5000+ 2.6 GHz Brisbanes that supposedly launched back then? Not at newegg, that's for sure. Whatever problems they are having can only be multiplied on a die more than twice as large.
May 23, 2007 6:43:44 PM

Quote:
Ok... I'm hearing that it was just a driver issue...


:lol:  ( j/k couldn't resist)
I was going to make that joke in another Barcelona thread, but I was afraid I would get flammed. :lol: 
May 23, 2007 7:02:44 PM

1) - I can't say. I can say that sales didn't go down as much as some would like to believe AMDs big mistake was price cutting They did cut X2 prices by > 50%.
2) - If they can't get through the 2 quarters to get this out what's the point.

3) - I don't think they're hidng it. I think they for one do't want Intel to know too soon and second they probably have work to do on Vista and Longhorn drivers.

Yes, I wish they would just come out with it but I'd rather see volume in Jul/Aug than benchmarks now.

All of that means I have no frigging indea.
May 23, 2007 7:54:39 PM

Dats bad

Tigerton breaking thru 10k?
May 23, 2007 8:01:13 PM

Quote:
Didn't we see things like this before where there was a bunch of tit-for-tat benchmarks and both Intel and AMDbested each other and then claimed that the other was not running an "optimized" system with outdated drivers? Yeah, I thought we had...

Well, here we go again! :roll:


Yeah, cause that's almost clinically impossible. Barcelona CAN'T be slower than Opteron.
Just wierd.
The problem is that actually, the only benchmark we have shows barcelona performs more or less like a K8. The only thing that can turn our assumption upside down, is that POV ray recognizes at most 8 cores, so that on both of those test systems, 8 cores were used, and, at the same frequency, would be pretty amazing if the K10 performed 2X better than the K8 core, but then again, I don't know, I'll have alook at POV-Ray's page to see how it works.
May 23, 2007 8:46:45 PM

AMD's official response....


Quote:
The objective of our demo was to show performance scaling from our current dual-core processors to our upcoming quad-core processors within the same thermal envelope and drive home the point through a real-world demonstration that customers could expect to see 2x the performance without an increase in power consumption.
At some point before our launch you can plan on us showing a demo of our parts vs. Intel's high-performance processors.
One thing to note, the system we showed, while it was a 4P, it was running only 6GB of memory.


http://www.uberpulse.com/us/2007/05/amds_barcelona_lack...
May 23, 2007 10:43:36 PM

He only said what we already knew, so I am not much convinced, especially after seeing in what state he was:
!