May 21, 2007 3:57 PM PDT

AMD shows sunny side of Barcelona

SAN FRANCISCO--Advanced Micro Devices is ready to start fighting back against Intel with silicon, but Intel has an answer in the waiting.

AMD demonstrated its Barcelona quad-core server chip for reporters and analysts here Monday, comparing its performance to one of AMD's dual-core Opteron processors. This marked the first time AMD shared Barcelona performance information with anyone outside of its server partners and internal folks, said Randy Allen, corporate vice president for server and workstations at AMD.

The company is on track to ship Barcelona in production volumes sometime around the middle of this year, Allen said. Server vendors already have samples of the quad-core chip, and systems are expected to start arriving in the third quarter, he said.

Barcelona represents AMD's current hope for returning to profitability by stabilizing its server processor prices. The company has been forced to dramatically cut the prices of its dual-core Opteron processors to compete with Intel's quad-core server processors, which have been on the market since last November.

The hope is that Barcelona's design, in which four processing cores rest on the same piece of silicon, delivers enough of a performance boost over Intel's quad-core Xeon chip to once again attract demanding server buyers. Intel chose to put two dual-core chips into a single package for its first quad-core processors in order to get out in the market well ahead of AMD; AMD contends that's an inelegant design that doesn't solve Intel's problems with memory bandwidth.

Still, Intel has quad-core chips now for those who want them, while AMD doesn't. Until then, AMD is talking up the performance of Barcelona as it works to get the chips ready for prime time.

The company ran a demo comparing the performance of two four-socket servers, one using the quad-core Barcelona chip and one using a dual-core Opteron chip. The demo measured the performance of the chips on an imaging benchmark called POV-Ray, and as you might expect, the quad-core chip finished its task quicker than the dual-core chip.

The quad-core chip processed about 4,000 pixels per second in rendering the image, while the dual-core chip could only hit around 2,000 pixels per second. Allen said Barcelona was not running at the fastest clock speed that will be available at launch, although he declined to specify the speeds that will be available.

AMD did not test Barcelona's performance against one of Intel's quad-core Clovertown Xeon processors, but Allen said Barcelona "will be the highest-performing x86 chip out there. It will blow away Clovertown."

But soon after AMD's Barcelona arrives, Intel will launch its Penryn processors, which are expected to be a significant improvement once again over the current generation. Allen said he couldn't project Barcelona's performance against Penryn, since Intel hasn't published any benchmark results based on the forthcoming chip.

Intel did announce several performance results for Penryn chips at its Intel Developer Forum in Beijing last month, but those were preliminary results that have not been published by a benchmarking organization, an Intel representative said.

See more CNET content tagged:
Barcelona, quad-core, AMD Opteron, AMD, dual-core processor

15 comments

Join the conversation!
Add your comment
When will willing accomplice hype end?
"But soon after AMD's Barcelona arrives, Intel will launch its Penryn processors, which are expected to be a significant improvement"

Define "significant". A die shrink does not classify as a significant improvement. And a die shrink does not guarantee an improvement in performance. All Tom has to do is remember when people were talking the same hype about "Prescott", when the Prescott shrink to 90 nm was going to be SO MUCH better. Yeah, it was a much better improvement if you want it to heat your house, but that was about it. And how can SSE4 be a significant improvement when there isn't software optimized for it yet? So why is it "expected to be a significant improvement." No reason at all, other than just because this is Intel and they say it will be, which they are not that I am aware of. All they are saying it is just a die shrink and additional instructions.

I'm just tired of all these wild assumptions about Intel that isn't founded on anything. If it was founded on anything, it would be founded on the 90 nm Pentium 4 Prescott die shrink which was a massive failure when everyone "expected it to be a significant improvement" as well. So history contradicts what he is saying when Intel performs a die shrink, alone, of an existing core architecture. Another example of unfounded hype was when Intel first came out with the Pentium 4. The same type of hype was coming out about that, similar to the Prescott release. "Oh, watch out, with the Pentium 4, they could hit 10 ghz!" I remember reading. NOT! It didn't even hit 4.

I wonder when people as major news and review sites are going to stop being willing accomplices, being nothing but intel hype machines by making these unfounded wild speculations, especially given the past two examples I mentioned above, that they are WELL aware of and were completely wrong. This is not a major change at all, just a die shrink. No architectural change, nothing, but a die shrink and additional instructions. That is IT. Core 2 IS what I would call a significant improvement and it was a significant improvement over the Pentium 4 netburst architecture, but this, the Penryn,is not. And can't be predicted to be without some kind of benching numbers. Which there are none.
Posted by redraider89 (9 comments )
Reply Link Flag
Soooooo
you dont like Intel? Enough to write a book about how the Pentium 4 die shrink was a faliure:)

The pentium 4 was a failed path of technology. Using a build process that had massive current leakage problems that led to heat problems. Shrinking the die was the only way to increase the speed but it was still a failed path of technology.....and still had heat problems.

Core 2 is NOT a failed path of technology. Shrinking the die, will make it only better. If they dont do anything but shrink the die, not even raise the clock speed it will make the chip even cooler, require less power and allow them to make many more CPU's with same wafer....and pass along the savings to us. So for something a like a blade servr....this will be really great news as now they use less power and run cooler.

Of course they will also be able to raise the clock speed a get a faster CPU with out rasing heat or power requirements over the .65nm Core 2 cpus.

Then there could be some enhancements to increase performance even more.

In the end your passion over a piece of silicon amuses me:)
Posted by Lindy01 (443 comments )
Link Flag
Heh... kind of sad
I am finding myself agreeing with Lindy mate, you cannot seriously get that worked up over a friggen computer chip, doing so makes you look like you have no argument, which in this case you really didn't.

Everyone has their favorite chip maker, maybe I prefer Intel for the sheer fact they have lasted me and worked well, but it does not mean id talk crap about AMD, they keep Intel working which means I get a better processor one way or another and that for me rocks.

As far as the whole die shrinking, Lindy said it well enough and I am quite happy to agree. Your passion over it all really does seem to make me smile, maybe you work for AMD? I just cannot see any normal person acting like that for a stupid processor making company. When times change, roll with it or look like a fool.
Posted by Rimer (24 comments )
Link Flag
Intel said Penryn 40 percent improvement over C2D
At their Beijing IDF, Intel projected that Penryn would be up to 40 percent better than the 65nm Conroe/Merom cores for certain tasks, like gaming and video encoding. Sure, those are just projections, but I'd say 40 percent counts as "significant," assuming they can deliver it, of course.

<a class="jive-link-external" href="http://www.intel.com/pressroom/archive/releases/20070416comp_b.htm" target="_newWindow">http://www.intel.com/pressroom/archive/releases/20070416comp_b.htm</a>
Posted by Tom Krazit (436 comments )
Link Flag
And everyone believes what Intel says....
I'm just tired of Intel getting a pass for horrid engineering and design. They should have been derided like Nvidia was when they came out with the Geforce 5800, aka The Vacuum Cleaner, but they were still treated like they are the best x86 CPU producer on the planet even though it set records for thermal dissipation (while AMD was hounded over the heat of the K7 Athlons, no one cared if the Pentium 4 set records for room heating ability, way more than the k7 ever did, too). And Intel said, again, the Pentium 4 would hit 10 ghz. And you people probably believed them when they said that, didn't you. And they said the Prescott core, which, again, was just a die shrink was supposed to bring new levels of performance, ACCORDING TO INTEL. Did it? No. But it did bring new levels of heat dissipation, as stated above.

And there is so much ignorance out there. That is what bothers me too. The blind ignorance. The "If Intel says it, it's true" mentality. This type of ignorance is one thing that hurts AMD, which hurts you and me, the consumer. So TRUE customer choice is hindered by this ignorance. If Intel is so opportunistic as to sell their laptops below cost to keep a not for profit organization from selling AMD based laptops, and so opportunistic to PAY OEMs to not carry AMD products, don't you fool yourself into believing that Intel won't really ratchet up the prices if they don't have competition. AMD isn't that strong financially, but you and I need them to keep Intel in check. How do you think they got so big in the first place. Because they were a virtual monopoly for x86 CPUs during the early 80s, built by over charging the consumers for their processors.

And the hypocracy is really thick with Intel, and alot of Intel users, the heat issue mentioned above is a prime example. And the giddiness of the tech press, it's like they have been Intel exiles over the past 4 years, when the Athlon 64 was ruling the x86 CPU roost. But now they can come home and crow from the roof tops, over amplifying every miniscule detail of everything Intel issues.

Then there is the misleading ad campaigns, designed to fool the masses. For example, "Dual Core. Do More". Do more than what? Do more than the competition because they are the only ones who has a dual core processor? Like they were the first to have an x86 dual core? It's pure misinformation. The fact is, you could do more with AMD before Intel because AMD had the first commercially available x86 dual core processor. And the only reason Intel has a dual core processor, WITH 64 bit instructions, IN THE FIRST PLACE is because they didn't like AMD beating them to having a dual core processor and the 64 bit instructions, the EMT64 instructions, and decided to do a "Me TOO!" on both of those features. I know people are thinking that Intel is the only processor with dual core because I saw a post where a guy said that AMD isn't as good because they don't have a dual core processor when they are the main reason we have them now. So, with every misinformation campaign, and every hype press release that Intel issues, the consumer, which is you and me, are harmed. So sure I get mad because I don't want to see Intel's main competitor go out of business because of propaganda and misinformation, and hype and then be stuck with over priced processors. I think that is worth getting worked up about.
Posted by redraider89 (9 comments )
Link Flag
I like AMD
I'm glad that AMD is still around (just!) It's kind of like the Apple-MS relationship: AMD is keeping Intel on its toes which benefits us all. Plus I always like an underdog.

I'm just disappointed AMD didn't come out with a quad-core earlier because I had to buy Intel instead.
Posted by a85 (104 comments )
Reply Link Flag
I agree with this report.
Wheres the beef?

Where are these chip on store shelves?
Nothing on the east coast for OEM retailers.
Posted by inachu (963 comments )
Reply Link Flag
I agree with this report.
Wheres the beef?

Where are these chip on store shelves?
Nothing on the east coast for OEM retailers.
Posted by inachu (963 comments )
Reply Link Flag
Where was the Core 2 in March of 2006
Are you sure this isn't hypocracy because you weren't asking "where's the beef" when Intel was leaking information about the Core 2 were you. You weren't saying "but I don't see the Core 2 on OEM shelves" were you. But if AMD starts releasing early information about their upcoming products, you say, "Where are they and why aren't they on the store shelves?" Sounds like a double standard to me.
Posted by redraider89 (9 comments )
Link Flag
Hey AMD! Stop Talking and Deliver!
I like AMD but they are currently at a disadvantage in the single and dual CPU systems both at the desktop and server. AMD still has the 3+ way CPU server market to themselves as far as performance goes but stop talking and deliver the goods already.
Being an AMD supporter I am afraid that they are taking way too much time to deliver their products. They are giving Intel the time to answer what ever AMD is coming out with so it is going to be back to AMD sitting in the back seat marketing their processors as "cheaper alternatives" rather than better technology.

AMD you had better get off you duffs and get some product out there NOW!
Posted by fred dunn (793 comments )
Reply Link Flag
I agree with this...
I don't agree that they shouldn't release information about their upcoming products, but I do agree that they got too comfortable and underestimated Intel's determination to get out of the number 2 spot. First the Athlon whipped the Pentium 3, so they came out with the Pentium 4. Then AMD came out with the Athlon 64, which at the minimum had the same performance, and most of the time had better performance, but did it with less energy and at lower frequencies. So Intel is determined now not to be beat again. AMD needs to develop the same drive and determination and not rest on past laurels.
Posted by redraider89 (9 comments )
Link Flag
Seems unlikely
Everything I've seen from OEMs and other reports puts AMD Barcelona launch in November, and high end in December. 2007 is the year of Core 2.
Posted by jprior (3 comments )
Reply Link Flag
Really now...
If you really want to tout performance, why didn't you perform the benchmark against and Intel Quad Core Processor?
Posted by enklein (1 comment )
Reply Link Flag
 

Join the conversation

Add your comment

The posting of advertisements, profanity, or personal attacks is prohibited. Click here to review our Terms of Use.

What's Hot

Discussions

Shared

RSS Feeds

Add headlines from CNET News to your homepage or feedreader.