Aceshardware

dougSF30 · Joined: 18 Sep 2007 Posts: 52

http://www.techreport.com/articles.x/13224/1

Nice clock/clock gains, particularly on the bandwidth-hungry apps, as expected.

(By contrast, the Anandtech article on the same topic is somewhat... nonstandard, with a strange benchmark selection.

http://www.anandtech.com/printarticle.aspx?i=3099

)

EDIT: And here is the Inq article back up:

http://www.theinquirer.net/?article=42423

...

Conclusion:

Should AMD be truly, madly, deeply concerned about Intel's new baby? In my mind, yes it should. It has a clear overall CPU/FPU speed advantage, fixing some loopholes where AMD beat the Clovertown before, combined with vastly improved memory throughput to close the last critical gap.

If AMD manages to pull the (possible new stepping) miracle and get, say, a 2.8GHz Barcelona out of the door in measurable volumes by year-end, that would be a coup de grace and a welcome event to all of us. Whoa, AMD will stay alive, no Intel monopoly for now! The hacks, ANALysts, PRs and all the other gang of that ilk can keep their jobs, as there will be more things to write about.

Humour - or relief - aside, I believe, based on the Harpertown's initial results, AMD needs a 2.8GHz Barcelona part to have a strong, balanced competitive position against a 3.2GHz Harpertown.

With benchmarks we saw at friends' places in Taipei, Barcelona is still slightly faster clock-for-clock in a few benchmark types, whether it is SPECfp or some database and non-SSE3/4 compression routines, as well as memory benchmarks. However, the current Oct/Nov shipment expectations, with 2.2GHz Barcelona vs 3.2GHz Harpertown, more than negate Barca's slight per-clock advantage.

On top of all this, hidden sources' silent whispers say that Intel may be able to go towards 4GHz on the Penryn generation real soon now - as in, by year-end - even with air cooling if they wish, after all this is the 45 nm process tuning test baby prior to the Nehalem's expected ~ Computex 2008 launch.

So, Intel could reasonably quickly shoot out a 3.6GHz or even higher speed TDP-busting but still production-grade Harpertown part if need arises. Will the same scenario repeat on the desktop? We wish AMD best of luck - they will need it against Chipzilla's new 45 nm weaponry.[/i]

TacoBell · Joined: 17 Aug 2007 Posts: 287

Anand for some reason did a CCP of their Barcelona article.

Tvar' · Joined: 27 Jul 2007 Posts: 24

From the Anandtech article:

Quote:

There is of course one area where AMD still does have an advantage: 4-way and higher server configurations, where their Direct Connect topology has some distinct advantages that may not be overcome for quite some time

Methinks Nehalem is becoming an elephant in the livingroom for AMD - it's the big looming problem that no one wants to discuss. I find it hard to interpret this wording from the Anandtech article as anything but intentional and needless Uncertainty spreading (the U from FUD). If they were being more objective, they would have just said "advantages that will not be overcome until Nehalem's release, currently scheduled for H2'2008".

I suppose the editors at Anandtech need to appease AMD, and after a dismal showing by Barcelona in this strange benchmark suite, I guess subtle FUD spreading was the best they could do.

Seriously, though - what advantages will Barcelona have after Nehalem is released? Any guesses?

Alberto · Joined: 04 Sep 2007 Posts: 111 Location: Italy

Anandtech article doesn't has a common sense. He is focused on two database benches without a single real world application.
It is an "academic" article with "academic" result, this is very funny:

"We estimate that a 2.0GHz Barcelona part should roughly be on par with a 2.66GHz Harpertown part"

LOL!! pair in what? in theoretical scalability? Yes in theory ;-) but what about "on the road" performance ?

We all know that these conclusions are not true with common sw even considering the old Clovertown on the old plataform. Anand needs of better collaborators.

Alberto.

EaS · Joined: 09 Sep 2007 Posts: 8

Two more reviews:

http://www.hardware.info/nl-NL/articles/amdnZWppZGSa/Nieuwe_quadcore_server_CPUs_AMD_Barcelona_vs_Intel_Harpertown/
(English version)

http://www.tecchannel.de/index.cfm?pid=212&pk=1729228

TacoBell · Joined: 17 Aug 2007 Posts: 287

Number of Harpertown reviews: >=6
Number of Barcelona reviews: 2 (incl 1 incomplete)

Wierdo · Joined: 27 Jun 2007 Posts: 184

Alberto wrote:

Anandtech article doesn't has a common sense. He is focused on two database benches without a single real world application.
It is an "academic" article with "academic" result, this is very funny:

"We estimate that a 2.0GHz Barcelona part should roughly be on par with a 2.66GHz Harpertown part"

LOL!! pair in what? in theoretical scalability? Yes in theory ;-) but what about "on the road" performance ?

We all know that these conclusions are not true with common sw even considering the old Clovertown on the old plataform. Anand needs of better collaborators.

Alberto.

Off-topic, but Alberto either you're not the same person, or your English writing skills are much better than they used to be. Good job on that :D

thibs · Joined: 03 Aug 2007 Posts: 119

Wierdo wrote:

Alberto wrote:

Anandtech article doesn't has a common sense. He is focused on two database benches without a single real world application.
It is an "academic" article with "academic" result, this is very funny:

"We estimate that a 2.0GHz Barcelona part should roughly be on par with a 2.66GHz Harpertown part"

LOL!! pair in what? in theoretical scalability? Yes in theory ;-) but what about "on the road" performance ?

We all know that these conclusions are not true with common sw even considering the old Clovertown on the old plataform. Anand needs of better collaborators.

Alberto.

Off-topic, but Alberto either you're not the same person, or your English writing skills are much better than they used to be. Good job on that :D

You're joking, right? :lol:

thibs · Joined: 03 Aug 2007 Posts: 119

TacoBell wrote:

Number of Harpertown reviews: >=6
Number of Barcelona reviews: 2 (incl 1 incomplete)

Your point is ?

TacoBell · Joined: 17 Aug 2007 Posts: 287

thibs wrote:

TacoBell wrote:

Number of Harpertown reviews: >=6
Number of Barcelona reviews: 2 (incl 1 incomplete)

Your point is ?

That something smells when a product still 2 months from launch has broader availability than one more than a week old.

Although Barc now goes to 2.5 and Harpertown to >=7,

http://www.hardware.info/en-UK/articles/amdnZWppZGWa/New_quad_core_server_CPUs_AMD_Barcelona_vs_Intel_Harpertown/1

2GHz Barc using a Dell server getting mostly stomped (losing on a per-cycle basis by 10-20%) except in FP rate base and a couple of int_rate_base sub tests.

thibs · Joined: 03 Aug 2007 Posts: 119

Because number of reviews of any product are indicative of anything (except marketting department quality) now ?

That'd be a first.

TacoBell · Joined: 17 Aug 2007 Posts: 287

thibs wrote:

Because number of reviews of any product are indicative of anything (except marketting department quality) now ?

That'd be a first.

When a product is 'released' it is indicative of availability and possibly or some deep problems (for example the PIII-1.13 failure to launch which IIRC had 2). When there are 6-10 reviews on day-0 it indicates that availability is likely to be good in the short term and that the HW is not overly burdened with problems.

Compare Barc to the FX7x launch for some AMD-to-AMD comparrisons.

thibs · Joined: 03 Aug 2007 Posts: 119

I don't think Barc introduction is in anything comparable to a FX7x introduction.

The fact that Intel has a nice number of samples prior to lunch is... usual for Intel and IMO I'll be very surprised when Intel will announce such a product not being able to do this.

Also, remember that Intel wants to ridicule AMD as best as it can (sure I'd do the same) so they had to so this. They did it and good point for them. And then?

TacoBell · Joined: 17 Aug 2007 Posts: 287

Quote:

I don't think Barc introduction is in anything comparable to a FX7x introduction.

It's a product launch on a production node that plugs into existing infrastructure. Well, the FX7x didn't have the last one, but close. Pretty similar on the surface.

The outcome of the two launches has been night and day.

Quote:

The fact that Intel has a nice number of samples prior to lunch is... usual for Intel and IMO I'll be very surprised when Intel will announce such a product not being able to do this.

This because, at least recently, Intel hasn't been paper launching much if anything. I would say the same for AMD in recent memory, except for a few parts (35W 90nm A64s, some 65nm A64s BE-seres). But this is a clear case where AMD appears to be having some sort of launch problems, whatever they may be.

Quote:

Also, remember that Intel wants to ridicule AMD as best as it can (sure I'd do the same) so they had to so this. They did it and good point for them. And then?

You state this as if the opposite isn't true. The company that is doing the ridiculing is probably in a much better position than the company being ridiculed.

Even AMD's god and pony shows, the 3GHz Phenom demos, have been so flawed as to provide little positive value for them.

If AMD wanted a marketing stunt they should have made 6 or 10 3Ghz Phenom system and sent them around to the usual suspects, assuming they are stable. That they haven't makes me think that they wouldn't be.