Last message on previous page:
Intel is simply a monster when it comes to R&D and product portfolio. AMD is struggling to release a single product, where Intel can simply grab something "off the shelf" and release it.

I have a friend who works at the Intel fab in Arizona. He tells me that Intel has several advantages over AMD. The first, is that each Intel fab is consutructed exactly the same, which makes upgrading and refining their processes much easier and faster. If they find that something works really well in a fab, they can quickly modify the other fabs to keep pace.
The second, is that at any given time, Intel has at least three new generations of product refined and ready to hit market within 3-6 months if they so desire to do so. Intel sits on a lot of product, with no real reason to release anything unless they feel pushed to do so. This is why it is imperative that AMD survive and refine Phenom so that Intel is pushed to continue to release new products. Otherwise it will be the good ole' days of the Pentium 2, where the cost difference was immense for a difference of 33 mhz

But that begs the question...

Did they KNOW that B3 didn't fix it? Because as far as I remember they claimed B3 would fix it. How long have they known B3 wouldn't fix it?

Or is it yield issues.

No, *IF* that is the case, add 10 weeks to the projected Q1 release and you get Q2.

I have no idea if B3 fixed it or not. AMD will know by now, but only about now at the very earliest.

All I did was answer your question... IF the B3 stepping still has the TLB bug, THEN they need to do a respin of the silicon, which adds 10 weeks to whatever they projected. It fits with the Q1->Q2 delay, I suppose, but it's highly circumstantial, and I would not put much stock into it.

I know what you meant, my question is I'm wondering when they realized the B3 still had the bug (if it does).

I can tell you that your friend is selling you a fantasy.

The last thing a semiconductor company wants to do is to make a lot of product they aren't going to sell, or worse, let it sit on a shelf somewhere.

I can emphatically say there are not, nor have there ever been, "three generations" of product just waiting to be sold.

Intel has three generations of product is various stages of production / development, and by this I mean process generations -- 65nm, 45nm, and 32nm, with the latter contained to their research fab. Maybe you include 90nm in this to get your three generations being sold, but that's *really* old hat. And each of those generations are in different fabs.

As for product -- again, unless you're talking about Itanium / x86 / flash or CPU / chipset, I have no idea what your friend is talking about. Let's take the heater known as Cedar Mill/Prescott. If intel had other stuff to sell (ie, the P-M), don't you think they would have responded a lot faster? Sure, there are different "grades" of Core 2 going out now (Server/DT/Mobile) which are similar, but not quite identical. And then there's silverthorn -- which *is* something new... and the Nehalem and/or Dunnington, the foray into "more than two cores" for intel, but these aren't exactly sitting on the shelf. Most likely they are working through silicon respins to get things *ready* to sell.

About now would be the right time.

DELETED Is that all you're capable of???bashing AMD???You better pray and hope that AMD does not go under cause if it does we are all screwed.So instead of bashing the hell out of them,try offering some support for a company that has shown time and again that it can put out quality high end products that beat INTEL hands down.So what,is it ok for INTEL to sit on their prestigeous past like they did for a couple of years before they came out with C2D???AMD beat them before and they'll beat them again.I do agree that Hector needs to take a long look at his career and the way he does things and try to do things differently.The purchase of ATI was a good move on AMDs part.The technology on the graphics side is getting quite complicated and although it is a huge expenditure initialy,given enough time,it will prove very valuable to AMD as well as beneficial to the consumer.MERRY CHRISTMAS EVERYONE.

Dahak
M2N32-SLI DELUXE WE
X2 5600+ STOCK (2.8GHZ)
2X1GIG DDR2 800 IN DC MODE
TOUGHPOWER 850WATT PSU
EVGA 8800GT SUPERCLOCKED
SMILIDON RAIDMAX GAMING CASE
ACER 22IN WS LCD 1680X1050
250GIG HD/320GIG HD
G5 GAMING MOUSE
LOGITECH Z-5500 5.1 SURROUND SYSTEM
500WATS CONTINUOUS,1000 PEAK
WIN XP MCE SP2
3DMARK05 15,686
3DMARK06 10,588

The only design amd managed to get in front with was the K8 (A64/Opteron/X2 series) - The K5, K6 and K7 all came close, but no luck. On the up side they have always been cheaper are easier upgrades (until K8) - socket 7 got years of like added with the K6-2 and K6-3, socket a lived a long time.

As for "discarded P3 and mobile architectures" - P6/Pentium Pro -> Pentium 2 -> Pentium 3 -> Pentium M -> Core Duo -> Core 2 Duo - its not discarded, and Intel had it on there road maps a year before it came out, i was actually speaking of conroe almost a year before it came out here on these threads, and no one thought it would work and happen lol.

Dahak, on this sacred occasion allow me to wish you and yours all the blesings of the season, a very merry Christmas and a truly happy, joyous and prosperos new year.

DELETED Ever hear of survival of the fittest? AMD isn't some mutant baby with one leg and one eye you keep alive because of theological ethics. Sure, a monopoly by intel woudl be hell, but why should I or anyone else subsidize with our credit cards the sheer incompitence and utter stupidity of Hector Ruiz and his bunch of self serving execs who are only looking at their exit checks after they cut their makret cap to 1/3 and put their company in the toilet? DELETED
Best of the season, peace on earth, goodwill towards men... even AMD fanbois like you!

Hence why before the new 45nm core 2's were even release Intel had samples of Nehalem floating around and even demonstrating them, this is well before even AMD had benchmarks for there new cpu, and even so before AMD even had working 45nm samples working (prolly still don't even today).

Intel Timna - altho not released had some interesting concepts, YEARS eariler (http://en.wikipedia.org/wiki/Intel_Timna) - 1990???? So much for AMD's radical fusion cpu design. Pfffftttttt

Nehalem is neither refined nor ready to sell. If you think the response to the Barcy was bad, if intel released Nehalem today, it would be a bloodbath. What was demonstated were some carefully chosen demos using A0 silicon. A feat, to be sure, but that does not a saleable product make. From a demo like that, you tack on 9 months at the earliest before that product is ready.

As for why they would demo it then? That's sheer marketing.

And AMD does NOT have 45nm samples working. If they were, you would hear the press releases. Immediately. Not "we're gonna have them soon", not "we're confident they will boot", but "WOW!! Lookit investors! Here's task manager on our 45nm chip!!!"

That is, unless you buy into the "Hector wants a golden parachute" theory.

Nehalem is 12 months away from sale, and when it comes out there is a resonable expictation that it will be ready for prime time, unlike that disaster that K10 turned out to be.

and yes, I subsribe to the Hector's Golden Parachute Theory. He's done it before, remember?

hence why i said samples?