Sign in with
Sign up | Sign in
Your question

Barcelona benches are not worth jack without facts...

Last response: in CPUs
Share
June 3, 2007 6:28:13 AM

http://www.overclockers.com/tips01162/
We need the tech facts and benchmarks not fanboi's empty ego's to believe it... and if you don't think THG and Fz have stooped lower, wait and see if management keeps meeting AMD's margins and ethics.
June 3, 2007 6:56:05 AM

Choir preachin'?

Who's FZ?

And what does THG have to do with AMD's margins and ethics? I think they accept adverts from both camps...

Just playing devil's advocate here, but if Jack has truly left, someone has to do it.
June 3, 2007 7:01:07 AM

Quote:
Choir preachin'?

Who's FZ?

And what does THG have to do with AMD's margins and ethics? I think they accept adverts from both camps...

Just playing devil's advocate here, but if Jack has truly left, someone has to do it.


Fz? I think that's Fudzilla, but I could be wrong.
June 3, 2007 7:28:02 AM

Oh... Fudzilla. Richpls, you might have also mentioned the Inq.

I agree, what is now needed is 3rd party benchies of both camps CPU's, using more realistic tests than just Specint. It would appear that in server apps the Barcelona will have some improvement over the Opty series, which now has a rather large advantage over the Xeons at least in the Specint benchmarks. But some real-world benches are needed... Server, maybe some Oracle (or other DB) benches?

Desktop... There are many benchmarks to pick from. Just bring them on. This speculation is not helping corporate buyers decide on next gen servers or desktops, and systems are always needing replacing.

My 2p.
June 3, 2007 7:50:54 AM

I'm not saying that these are inaccurate.. but.. I am still waiting...
June 3, 2007 8:01:29 AM

I think that there are enough benchmarks (based on specint) of opty's vs xeons to not doubt that these benchmarks are fairly accurate. Do they tell the whole story? No. Do they predict Penryn vs. Barcelona? I don't think so. So to draw the conclusions that this particular article drew, well... That's 'drawing' a very long bow in my opinion.
June 3, 2007 8:49:31 AM

Crap Crap and more Crap, if you want to compare something run they both at the same speed.
Dont just show the benches the benefit K8 and don't do so hot fot Intel. Come on guys, like is been said, FUD.
June 3, 2007 9:49:21 AM

This sounds just like the lead up to the core 2 launch, all the amd fans saying the benchmakrs are wrong, only for the core 2 to will when it comes out, now it's the intel fans saying amd can't beat the core 2, will history repeate itself? if amd wins back the performance crowns thats not saying much for intel when it took them 3 years to get it back after the amd64 came out
a b à CPUs
June 3, 2007 9:58:23 AM

Was the subject title a play on words with "not worth jack"?

*Takes a minute's silence in commemoration*
June 3, 2007 10:05:12 AM

I shall patiently wait. My old opty can stutter along just fine for now, as can my sli'd 6800 ultras. As to vendors, I buy from whomever has the best when I need it. Oblivion is quite nice in Vista, maybe I'll need to upgrade after (and if) Crysis comes out.
June 3, 2007 10:08:05 AM

Patiently wait.... Good things come to those that wait.
June 3, 2007 12:49:00 PM

Quote:
Was the subject title a play on words with "not worth jack"?

*Takes a minute's silence in commemoration*


:cry: 
June 3, 2007 5:56:23 PM

Quote:
This sounds just like the lead up to the core 2 launch, all the amd fans saying the benchmakrs are wrong, only for the core 2 to will when it comes out, now it's the intel fans saying amd can't beat the core 2, will history repeate itself? if amd wins back the performance crowns thats not saying much for intel when it took them 3 years to get it back after the amd64 came out


There is a huge difference between this and the launch of C2D. With C2D there were dozens of 3rd party benchmarks, tests and analysis. With K10, to date, theres been no 3rd party testing. The comparison Rich links to is an estimation based on very soft information, which means as Rich points out, it is worthless.

The big arguments the fanboys tossed with the tests of C2D was that they were "cooked" or falsified benchmarks. These turned out to be blantently false accusations. With K10, there have been no benchmarks to be falsified or "cooked".
June 3, 2007 7:00:28 PM

Quote:
Who cares about K10 benchmarks? Teh only thing that matters is if it fast enough. If it is fast enough to boot Windows then it's a win for AMD.

ALL HAIL THE DUOPOLY


I have to disagree with you there Baron. I really think we need strong independent benchmarks.
June 4, 2007 1:51:21 AM

I read that article, and honestly im gonna take it with a grain of salt now, i went through some old articles on overclockers and couldnt find one positive thing said about amd or apple for that matter. I'll beleive it when its here or on anandtech, or xbit labs.
June 4, 2007 1:55:23 AM

Quote:
Was the subject title a play on words with "not worth jack"?

*Takes a minute's silence in commemoration*


:cry: Dude jack is a big boy, this is a democratic forum, he should accept that, i like jumping jack,and i like baron matrix, hell i even like his doppelganger matrix baron, but just let him ofr them cool off, he'll be back.
June 4, 2007 2:05:20 AM

If you flashback to one year ago and did a comprehensive look at benchmarks of the FX-62 (a Chip we now call an x2-5600 @ $19x) and compared it to an Intel EE 965 you could indeed find that out of abut 60 benchmarks floating around, the Netburster did indeed win a half dozen or so of them.

Intel could (and occasionally did) run ads showing these selected benchmarks and thus "proving" that the EE 965 was indeed the king of the hill.

Mark Twain noted that "there are lies, damned lies, and then there are statistics" and that is certainly true.

Now it seems AMD is trying to do the same....

That being said......

Opteron was, and remains, a damn good server ship. In the real world, it still uses less power than the Intels (due to cool and quiet which works much better than speedstep) The Opteron due to hypertransport scales better than the Xeons, and Opteron deliberately traded some clock speed for deeper parallelism which works in server loads better than gaming and office appications

This was smart on the part of AMD.

Firstly, to be taken "seriously" AMD had to get into the server market, and secondly, I am sure AMD makes more on a single 8xxx series Opteront han they do on a 100 low end x2s and Seperons they schlock out to Dell and baergain basement prices.

I think Barcelona is agian aimed at the same server. There has been a blizard of benchmarks showing that in games and office productivy C2D smokes the Opterons/Athlons, and in large measure these are true, but given the remaining strength in Opteron in the server segment, I don't think Barcelona reclaiming a lead in the Server segment is outside the rhelm of possibility.
a b à CPUs
June 4, 2007 2:17:40 AM

Those benchmarks are 100% accurate. But you see there's more to it then you guys seem to realise.

Under this specific test AMD's Fastest K8 (Opteron 4P/8P) has a 30-35% lead over Intel's Fastest Xeon's (4P/8P). And Barcelona at 3GHz has a 40-45% lead over Intel's current fastest Xeon (2.66GHz Xeons vs. 3GHz Opterons).

So what we can conclude here is that Barcelona is only 5-10% faster then current K8's in this test and thus increasing a lead AMD already had.

This says nothing about desktop or other application performance.
June 4, 2007 2:42:20 AM

Quote:
Those benchmarks are 100% accurate. But you see there's more to it then you guys seem to realise.

Under this specific test AMD's Fastest K8 (Opteron 4P/8P) has a 30-35% lead over Intel's Fastest Xeon's (4P/8P). And Barcelona at 3GHz has a 40-45% lead over Intel's current fastest Xeon (2.66GHz Xeons vs. 3GHz Opterons).

So what we can conclude here is that Barcelona is only 5-10% faster then current K8's in this test and thus increasing a lead AMD already had.

This says nothing about desktop or other application performance.


Yo hooo. There arent any benchmarks of K10. If there were, they would just have done a direct comparison.
June 4, 2007 2:44:30 AM

A whole lot of speculation about speculation. Wake me up when Computex comes around.
June 4, 2007 4:04:41 AM

Quote:
Those benchmarks are 100% accurate. But you see there's more to it then you guys seem to realise.

Under this specific test AMD's Fastest K8 (Opteron 4P/8P) has a 30-35% lead over Intel's Fastest Xeon's (4P/8P). And Barcelona at 3GHz has a 40-45% lead over Intel's current fastest Xeon (2.66GHz Xeons vs. 3GHz Opterons).

So what we can conclude here is that Barcelona is only 5-10% faster then current K8's in this test and thus increasing a lead AMD already had.

This says nothing about desktop or other application performance.


Yo hooo. There arent any benchmarks of K10. If there were, they would just have done a direct comparison.



My mind is going, yes, I can definitely feel it. Won't you stop, please stop. I really think you should sit down, take a stress pill, and think about this. You know I continue to have the highest enthusiasm for the mission.

Daisy... daisy... wherrrre arrre yooooooou gooooiiiiinnnng tooooooo.....
June 4, 2007 4:23:42 AM

Quote:
Those benchmarks are 100% accurate. But you see there's more to it then you guys seem to realise.

Under this specific test AMD's Fastest K8 (Opteron 4P/8P) has a 30-35% lead over Intel's Fastest Xeon's (4P/8P). And Barcelona at 3GHz has a 40-45% lead over Intel's current fastest Xeon (2.66GHz Xeons vs. 3GHz Opterons).

So what we can conclude here is that Barcelona is only 5-10% faster then current K8's in this test and thus increasing a lead AMD already had.

This says nothing about desktop or other application performance.


Yo hooo. There arent any benchmarks of K10. If there were, they would just have done a direct comparison.



My mind is going, yes, I can definitely feel it. Won't you stop, please stop. I really think you should sit down, take a stress pill, and think about this. You know I continue to have the highest enthusiasm for the mission.

Daisy... daisy... wherrrre arrre yooooooou gooooiiiiinnnng tooooooo.....

:lol:  :lol:  :lol:  :lol: 
June 4, 2007 6:22:34 AM

Quote:
A whole lot of speculation about speculation. Wake me up when Computex comes around.


Amen, Brother.
June 4, 2007 9:17:43 AM

Quote:
you also have to believe that current K8s are better, clock-for-clock, than C2Ds.


anyone believes in this ??? i certainly dont.
!