Options
Hammers Getting Tossed?
Before we go for the jugular, we'll note that AMD seems to want to pump socket 754 for a few quarters, then suddenly turn around and dump it. This ought to make those interested now in socket 754 systems think twice about it.
But that's not what is most important here.
What is important is that AMD doesn't plan to make many Hammers over the next year, period. Even a year from now, desktop Hammer production will be less than 20% of a decent production quarter for Dresden today.
[link=http://www.overclockers.com/tips00552/]Overclockers.com On the Hammer[/link] - Submitted by madmat
But that's not what is most important here.
What is important is that AMD doesn't plan to make many Hammers over the next year, period. Even a year from now, desktop Hammer production will be less than 20% of a decent production quarter for Dresden today.
[link=http://www.overclockers.com/tips00552/]Overclockers.com On the Hammer[/link] - Submitted by madmat
0
Comments
I've said it before and I'll say it again: The home market and the "enthusiast" market are drops in the ocean. This market that we love so much is way too "niche" to make any real money. When AMD gets into the corporate datacenter, that's when the real money gets made. If you don't think AMD is gunning for the enterprise sector, you're smokin' something funny.
I think AMD has to build a grass roots reputation as well for long term success. I really wonder sometimes how much of AMD's marketing does come from us enthusiasts; their isn't much marketing in other quarters, is there?
Right, furthermore, Dresden fab does not mfr just Opterons, they mfr what used to be called ClawHammers (pre-marketting) also. Its not so much that it is only 20% producing, as that the mix is partly Athlon FX and other chips, some of which do not need the 754 socket to work. AND, AMD is in the government research sector (Scandia Labs and other research centers have HUGE multitiered Opteron clusters), the game dev sector (Industrial Light and Magic's rendering CLUSTERS, for example), and Microsoft's High Tech center does not use just Intel neither. They are also in the educational sectors in labs that do physics research, for example, or teach advanced 3D CAD\CAM. Some of the Scandinavian collider and supercollider physics research centers have Opteron clusters going also.
BUT, most smaller business that is not doing stats mining or advertising graphics work is not needing a highly tuned VERY HIGH SPEED FPU multipiping, it is using more ALU calcs. For that, Intel has a good strategy so far. The stats mining systems SHOULD use Opterons, and in fact the distro folks for Linux have seen that light, they are releasing and selling Opteron-compliant distros and updating them. Oh, IBM is wselling Opteron based server clusters if you want them, also, or leasing time on same if you do not need always-on Opteron clusters or need to lease peak load compensation time on their servers.
Nah, Hammer is name-morphing, not dying. Socket 754 might not be needed for all of that, but IMHO, it will be superceded, and not flatly dropped. Socket 754 life and Hammer life do NOT equate as both or none. In several current prod "Hammers" and perhaps in more to come, PROBABLY more to come-- on a new socket and package. AMD also has a history of grabbing niches Intel does not dominate, then leveraging them to get a spread to other sectors.
John D.
Overpaid bunch of pansies is what they are...
This is a hammer:
What do you mean "BS"? It's a sledgehammer. Although, in point of fact, we're talking about clawhammers here, not sledgehammers...
Everyone knew socket 754 was getting dumped, why is this guy surprised they're going to decrease production of socket 754 chips, has he been antarctica for the last 6 months?
Amen brother.
I second this, based on personal experience. When my department head suggested an AMD-based system for one of our rendering machines, the IT dept guys just went: " What? AMD? What's it good for? " If I was there, I'd have sermoned them about IPC's, FPU performance, pipelining, vis-a-vis code optimization, instruction sets, etc. But naaah, they'll never have a clue about these things anyway. I guess my point is that AMD will just always have to be a niche product for enthusiasts, highly appreciated by a small portion of the computing industry, yet mostly ignored by the vast majority of users who are sadly uninformed.
Interesting point you've raised. I thought about this more than once, and come to realize that our segment (enthusiasts) seems the only ones pushing the actual merits of going the AMD route on grounds of more tangible issues, ie price-performance ratio, and of course the fact that it is a product that actually works good sans extended instruction sets and software optimizations. The mainstream marketing efforts out there isn't obviously enough. Doing "AMD Me" is a little silly I think. It is never enough to educate Joe Sixpack, lots of times I wonder if its really worth the trouble at all trying to inform/educate the Joes. I have friends and officemates who are your typical Joe Sixpack, and I just gave up a while ago, most of them just don't get it.
You might like that article from MediaMan, fateshammer
A lot of "IT" managers suffer with dwindling or non-existent budgets trying to make failing systems that belong in yesterday make it through tomorrow. A lot of them do what they do very well but only that. The cannot think outside the box and their scope of foresight ends with the close of their textbook on their last day of school.
Some...not many...are enthusiasts who love to tinker and live on the bleeding edge like we do...but dwindling budgets restrict.
Regardless...articles like that follow rules of news and marketing;
- sound like you know what you are talking about.
- ride the wave don't make it.
- pick a controversial topic and you'll be noticed.
That's the problem with people on a soapbox. If you stay in the park...you will have to listen to them. It doesn't matter if you agree or not. It doesn't matter if they know what they are talking about...or just talking to attract attention.
That is a really insightful, witty and substantial article. I like it a lot. But I have to say that it is too optimistic. With all due respect to the author (hats off to you sire) and those who feel strongly for it, I tend to be realistically pessimistic when it comes to AMD waging war on the marketing front. I just don't see how it is able to sustain against Intel marketing-wise, notwithstanding the fine points outlined by the author. In the last couple of years I am of the notion that Intel has been transformed into nothing more than just a marketing juggernaut, and it has become lethal in this in the amount of slanted info and misinformation sown among the mass market. The idea of AMD facing this head-on is just financial suicide. And in a very predatory consumer market, its a sad fact that a very good product actually gets driven to the ground by the onerous weight of neiscience. I hope I don't get flamed for this. But I'd like for anyone to take me to task on what I stated here in an objective and civilized manner, that way I might be enlightened on some points that are not too obvious to me. This is a great forum, I'm finding out, a lot of wickedly sharp and knowledgable peeps, I'm glad I subscribed here. Peace.