AMD releasing Athlon-branded K10

The chip can be ordered already, see http://shg.dk/p/2115417, a danish internet shop. It is however not on stock yet. At a price of around 700 dkr I need to see a benchmark before I would call it a bargin.

Best regards, Thor

They we're most likely tested as quad cores with a 95watt TDP but two cores came off as bad so they just tagged them and stored them. It's probably cheaper to test once for tdp then retest quad cores that show as having two bad cores. They most likely do have a far lower tdp but it's not worth AMDs time or money to double check them for lil or no monetary gain

Do folks have something specific that these are quads with 2 cores disabled or is this just speculation? How dysfunctional does the 65nm process have to be for this to be enough volume? And how with 2 cores disabled, is there no TDP improvement?

And to all of the business 'experts' out there who think it is better to sell something then throw it away... that is true if it DOESN'T IMPACT the rest of your product lines! What does tri-core do to dual core pricing of the K8's? What does it due to people considering either a 'low end' quad or a 'high end' dual core.

I have no problems with a disable dual core (again if this is true... I have seen nothing to state that is what this is) as it is operating in a space that already exists. However, when you look at the volume of tri-cores that AMD has sold, weigh that against the potential impact on the dual core prices(which are probably >10-20X the volume sales). When you consider the profit margin on the tri-core chips, a mere few dollar per chip impact on dual core sales would wipe out any gains from "found" tri-core revenue. And as crazy as it sounds, if the lack of tri-cores enabled AMD to sell dual core chips for a few dollars more - it actually might be more profitable for them to throw them away!

The flaw in all of these "hey it's better to sell it, then throw it away" arguments from all of the wannabe MBA's commenting is it is based on the assumption that sales of a tri-core is operating in a vacuum and is having no other impact on other dual or quad core sales or pricing. For these dual cores it should be less of an issue, but it will likely put some pressure on K8 dual core prices that are in the nearby speed bins.

Hey come on guys! Intel is also doing this thing. 
Last time I work at intel Test module, they have a m/c that can 'bin' specific chips. They are programmed to segregate chips according to their speed. In this case AMD also segregated things according to cores then speed. Go it? THat makes sense, rather than scrapped the products and go out of business.

this isnt idiocy, its an example of what to do with parts that dont make the cut. if all four cores dont work, sell it as an x3. if only two work, you sell it as an x2. amd management has plenty of buisness sense, for example: selling x3's and x2's instead of keeping the price of the phenom x4 high to pay for the chips that get tossed...makes perfect sense to me.

So as the tricores are failed Phenoms with one disabled core, so do these parts look like failed quads with 2 broken cores. Instead to scrap them, try to sell them... 

if its 9950 X4 you seek, its $180, if you go to bottom of X2 line, Now mere 5200+ its less than $70. When Sales Lag, Niagria Falls. Took effect yesterday in europe.
drashek

...if these were simply Phenom Quads with two cores disabled, AMD probably collected all Phenoms with two faulty cores since the production of Phenom started, afterall it's still better to sell those crippled parts at a low price than to not sell them at all.

Just like the 8000 series Phenoms these new "Athlon X2" cpus will do a good job of lowing the lost revenue from bad 4 core processors. Or did no one think of that as a reason to put out "dual core phenom" cpus? 

One company's trash, is another consumer's treasure. Wonder if the other 2 cores will be fused off or if you will be able to tweak them into working.

I wonder if this isn't going to be a really nonfunctional quad core which is half disabled. It wouldn't much surprise me. It would also be a good explanation as to why dual cores took so long too. They were waiting to have enough crippled chips to sell in quantity at different speeds. 
Or they were waiting to have enough to sell as a 2.3 GHz unlocked part.

This shouldn't be 95Watt TDP for a 2.3GHz dual core (maybe 45W or 65W?) - you are telling me a quad core (K10,2.3GHz) has the same power bin as a dual core? Maybe this is a 3.3GHz part? (which would account for the higher wattage) I think x-bit labs must have gotten either speed or power wrong.

This also makes no sense from a business perspective - wasn;t AMD talking up a 'rapid conversion' from 65nm to 45nm - why make these parts and incur tapeout and mask costs for a part with less than 1 year of lifespan and will in all likelihood be produced on 45nm? If 45nm is indeed healthy why not just focus this on 45nm and keep cranking out K8 65nm dual cores until they get the 45nm capacity up to make these. It is amusing how AMD mgm't seems to have no business perspective / planning whatsoever. Perhaps they can release even more triple and quad core bins (which still account for probably less than 10% of the desktop volume) and further mess up their whole pricing tiers and drive the high volume dual core parts down even further.

No they'll just release a K10 dual core part right smack in the middle of the K8 bins, so they can make chips with bigger die area and lose even more money? The more you consider the insanity of this, the more it sounds like a mistake on power and/or speed.

AMD have been making a balls of their processor designations ever since they introduced Skt 754. Remember when they were selling 2 or 3 different versions of Athlon 64 2800's, 3000's, 3200's, on the same socket? And then another set of the same numbers on a different socket? Then did the same with the X2's?

Bunch of muppets.