AMD: No 'Bulldozer' Anytime Soon
AMD's senior vice president and chief marketing officer Nigel Dessau said in a blog post that the next-generation Bulldozer architecture will not be available any time soon.
The funny thing about Nigel Dessau's post is that he actually has an interview with himself. In some ways, it seems like some kind of April Fool's joke, but on the other hand, an online prank just doesn't seem like the proper business-like thing to do, especially for huge corporations such as Intel and AMD. No, Dessau decided to have a Q &A session with himself to answer a bunch of questions that have recently come his way, especially in light of Intel's recent release of the Nehalem processor.
However, rather than talk about upcoming AMD products, Dessau begins the blog post talking about flattery, that Intel's Nehalem (or "Opti-clone" as he calls it) is merely a replica of the AMD Opteron processor. He even goes on to quote a statement made by Intel's Patrick Gelsinger found an article published in The Wall Street Corner, a statement that Intel's Xeon processor 5500 series is the foundation for the next decade of innovation. "Well, I almost agree with that," Dessau concurred in his blog. "After nearly six years of telling customers that the AMD Opteron processor architecture was the wrong answer, this week our competitor has finally delivered “Nehalem” ― which some might call a copy, at least as far as the architecture is concerned."
Thus, he begins his self-interrogation with the following:
Interviewer:
Wow, you people must be really scared.
Me:
Nope.
Interviewer:
Nope? Ok, how about really, really scared?
Me:
Nope. Nope. No, really we’re not.
He goes on to explain the reasons why AMD feels so secure: the company has been in this boat before. Over the last ten years, both Intel and AMD have traded performance leadership between each other as if passing the Olympic flame--around six times in fact. As recently as 2006, Intel made similar claims as the latest quotation yanked from the Wall Street journal. "We heard some of these same statements from Intel then, about an eighty percent performance advantage and never losing another benchmark to AMD again with “Woodcrest.” That gap closed quickly," he said.
Interestingly enough, he brings up a good point in the blog: more than 90 percent of what AMD sells is not its fastest part. Dessau said that the market for the fastest part is always small, and especially small now during a plummeting economy. Although Intel may leapfrog in raw performance with the overhaul of their server architecture, Dessau said that Intel is also introducing an new learning curve and resource drain for an already cost-sensitive and "disruption-averse" IT environment. Current consumers want value, consolidation, and ways to save money. "With all of our competitor’s talk about memory bandwidth, they have ignored the market that cares the most about having a large memory footprint ― the 4P market," he said.
By the end of the blog, Dessau moved on to talk about the 6-core "Istanbul" processor, which is apparently on track for launch in the second half of 2009; Istanbul will be compatible with existing OEM platforms. Then in 2010, AMD will introduce the next-generation "Maranello" platform that will feature the 12-core "Magny-Cours" processor. "Maranello" will also serve as the platform for the "Bulldozer" architecture slated to debut in 2011.
"Nice job Intel, but value for money is what’s key in this market," Dessau concluded.
Here's my suggestion for a title: "AMD Unconcerned by Nehalem Release." Short, to the point, and most importantly accurate. Really, whoever edited this article and let it go live with that title should recheck their journalistic integrity.
What AMD really needs is to create brand awareness, like Apple. Make it somehow "cool and fashionable" to walk down the street with an AMD powered laptop, or a bunch of creative yound college students sitting around an AMD PhenomII desktop in their dorm, or a smart engineer working around an Opteron workstation, etc.
Of course, that's easy to say, but unless an asteroid with $10B of marketing budget magically (and softly) lands besides AMD's headquarters, grand-scale advertising is pretty hard to achieve for a cash-strapped company.
I disagree... the "mindless consumer" is going to head to your local Best Buy or other big box retailer to get their computer. So ultimately it's up to the OEMs to determine what the consumer purchases. Which in a performance per dollar aspect may just be an AMD.
What AMD really needs is to create brand awareness, like Apple. Make it somehow "cool and fashionable" to walk down the street with an AMD powered laptop, or a bunch of creative yound college students sitting around an AMD PhenomII desktop in their dorm, or a smart engineer working around an Opteron workstation, etc.
Of course, that's easy to say, but unless an asteroid with $10B of marketing budget magically (and softly) lands besides AMD's headquarters, grand-scale advertising is pretty hard to achieve for a cash-strapped company.
Bleeding edge doesn't make the money in the end, it's nice to hang your hat on but most people want performance/$ and that is where AMD shines again and again.
Here's my suggestion for a title: "AMD Unconcerned by Nehalem Release." Short, to the point, and most importantly accurate. Really, whoever edited this article and let it go live with that title should recheck their journalistic integrity.
I hope you dont talk like this in real life
Intel = better by making money
I want to comment that in order to be in the mindset of people you have to be leader. People likes to buy the top-dog brand. When AMD surpassed Intel's performance crown with Athlon over P3 and Athlon64 over P4, they managed to turn profitable, and this processors weren't cheap. They had many expensive (premium category) high-end models (many above Intel pricing) . Lets face it: If you want to sell in midrange you MUST have a strong position on High-End. When your next high-end generation arrives, your last one turns into the midrange, and so over...
I built the whole unit in a 1U chassis for under $3000... Good luck repeating that with Intel. In fact, I could probably keep it under $5000 and build an 8p AMD server. Intel just doesn't have the best options for the middle ground server market unless you just have a LOT of money to burn.
AMD is getting stronger and i want to be part of it.
Here's the reality. The i7 destroys the Phenom in virtually every benchmark, and is roughly the same size. The fallacy in all this is that AMD can make chips cost competitive with Intel, but in reality, they can make the same size chip a whole lot slower. It's not a good situation.
If the Phenom II were a lot smaller, and cheaper to make than the i7, I'd be very happy with AMD's lineup. It's not, it's roughly the same size, and BADLY underperforms it. Right now, Intel positions it high because AMD can't compete with Intel's previous generation, but there's no reason going forward that Intel will continue to position this processor at premium prices. They do because they can, not because they have to.
AMD desperately needs something new. What this man is saying is, our processors are as good as theirs when it doesn't matter. What kind of logic is that?
For servers, it's all over for AMD. The performance per watt is dreadfully against them, and it's going to be very difficult to maintain anything close to their current market share. Intel gave them a gift with the Pentium 4 architecture, and again with FB-DIMMs. With the i7, they're giving them a nightmare.
I know he sounds confident, but what else can he say? In reality, he's got to be seriously concerned about the i7. Anything else just doesn't make any sense.
Three years ago, it was true that AMD had conquered the low end. Two years ago it was starting to change. Before the Phenom II came out, I would have chosen Intel dual-cores above the low-end AMD's because I got better bang for the buck.
I haven't had occasion to build an AMD-based system for myself or for most clients for the past three years. My first-ever build was an AMD, so I've a soft spot in that sense, but when I compare benchmarks and clearly see Intel coming out on top for most of the setups I come up with, I go with what works best. Insinuating Intel has the high-end but AMD has the low-end is something that should be tossed out the window because things aren't like they were a few years ago. The same philosophies don't work.
Granted, the new Phenom II 940 is excellent and I'm actually going to finally build another AMD system again since it's such a great bargain and offers a number of perks over the low-end core i7 build I was considering, but Intel has been incredible in the low-end market and I have no idea why nobody seems to know what's going on event though all it takes is a bit of cross-referencing with prices and benchmarks/reviews.
Oh well. Stupidity rules, so enjoy.
Absolutely, they are a big company, so they should dress up nicely, talk about return-on-investment, shareholder-value and they should never ever do anything funny or geeky.
The average customer you refer to doesn't know that intel produces processors. The average customer responds with "windows xp" when he is asked which version of office he is using. The average customer buys a hp, a dell or other system. He doesn't know if it has an intel or amd cpu, or even how big his harddrive is. He just know that it can hold 132.000 mp3s, cause that's what the supermarked advertisement told him.