Replies to post #62889 on Advanced Micro Devices Inc (AMD)

[BUGGI1000]

@Rupert/Rink
Rupert, your mobile arguments are right. I have to agree, but
I wanted to spread my main thoughts. Every AMD mobile sale
is one sale, which Intel didn't get, so we could count it
twice. When we see this side of the coin, we shouldn't under-
estimate this sector, but I have big doubts, that this is
an easy play, but who said this will be otherwise. ;-) So lets
start ...

Rink, well said. Couldn't agree more. Yonah will not play a
big role in 1.half of 2006 at the desktop-side, just because
all the supply will go into the notebook-sector. I see it
later in smaller cases or MMC devices, but with these specs
OEM could today buy from AMD. We only have to look at this
picture, how much power the x2-3800 consumes under load.



Buy a manchester part now and get the specs which you will
get later with Yonah. Btw: in all future desktops design, it
will not matter, whether you need 30, 40 or 60W, when you
don't want an really small device. It does matter, that you
don't need 90 or 120W, like the most Intel desktop parts today
. So, the X2 would fit in all future case (much smaller)
designs, when!!! and thats a big when, when the OEM's want.
When we think about that topic, its clear, that in the 2nd
half of 2006 a few more Yonah-derivates will appear in desktop
-like devices (PC), BUT when we look at the hole volume numbers
it will be a small, really small part. So, Intel has to pump
Celerons, Celerons, Celerons and "old" PIV, PIV and PIV. Thats
their main device today and that will be still their main
device in the near future. Oh my god, did I say already, that
these parts are constrained, due to power consumption and no
real updates will appear? Did I say that already? ;-)
So its clear that at the desktop front - there are lying big,
no, HEAVY volume numbers, in front of us. Volume numbers,
which now are fitted with around 50% Celeron and 50% PIV's.
Just look at the Celeron - more than 10+ mio. units of these
crappy parts are sold every and every quarter - its a shame.
Its a shame to see people buying this outdated stuff. Did I
speak about the great chances? ;-) I think I should end,
otherwise ...

[pfosse]

I understand your argument very well about desktops but my main questions is about Intel's manufacturing and marketing capacity to fight the desktop space on price, thereby squeezing margins even further. AMD would be gaining more and more of less and less.


We haven't talked die size in a while, but there are several interesting things with die size. In the desktop space, AMD has a much smaller core and then uses a smaller cache on top of it to have a much smaller die size for the same or better performance. Intel has much better cache density, which partially makes up for their need for larger caches. Intel 300nm wafers also partially makes up for their die size disadvantage. Overall, AMD has the advantage here until Intel goes to 65nm. So from now till 1Q06 or so, AMD should be able to gain share pretty quickly unless world demand grows too fast. In the case demand is very strong, AMD should just cut down on cache size and make more chips (since competition isn't strong in a high demand environment).

When Intel is at 65nm and AMD is at 90nm, Intel has the advantage (if they have decent yields), but when AMD gets to 65nm, they get double advantages (65nm and 300nm) at the same time, so it is back to AMD's smaller cores and less need for cache. About this time (late 2006), Intel comes out with their new core and it looks like it is small like AMD's, but possibly more powerful (4 instructions at a time vs. 3) but still missing the integrated memory controller and scalability of hypertransport. Clearly AMD needs to update their core sometime in 2007 and not just go to dual core, quad core, 8 core, etc. We have seen nothing public on this. If they indicate no new core at the November Analyst day, I will start to worry.

[bobs10]

I wonder whether Intel would also survive better in the event of an economic slowdown which would weaken consumer demand for desktops? (I do understand the argument that Intel's cost structure is less well able to sustain lower prices than AMD's in an economic slowdown. But I'm not convinced. Intel has no debt and lots of cash. AMD has no free-cash and lots of debt).

Me...

That's not the way it works. Your right about INTC not being under any threat due to its' large cash hoard, but long before INTC faces liquidity problems the stockholders/BOD will start asking lots of very uncomfortable questions of INTC management. These questions will result in multiple reorganizations and management changes that will cause INTC all sorts of other problems while it sorts thing out.

Personally, I expect to see cracks turning into fissures about the time AMD takes 25% market share. Up to now AMD has essentially been playing a defensive game. Mainly because of capacity restrictions AMD has been relegated to just retaining existing market share in an expanding market. Next year the flood gates open and AMD goes into full court press, 24/7. That's when INTC is going to feel the real pain of AMD's boot up its' ass.

One thing I haven't seen talked about is the bear trap Hector has set for INTC. The timing of the suit does seem odd unless you think about it in a wider context. After all why not 2 years or even 5 years ago? It isn't like INTC just started these dirty tricks, this stuff has been going on ever since Hector was a pup. The Athlon debut just resulted in INTC flaunting its' arrogance and contempt of AMD in an out of control way.

To me the suit was initiated because of the expected settlement date, late 06. This date just happens to coincide with the time when fab36 will be reaching full 65nm capacity. What I expect from the settlement is an award of at least $2B to AMD and a ruling that INTC can no longer give kickbacks based on customers restricting AMD purchases. Outside of this I don't expect the courts to specify exactly what INTC has to do so I expect a lot of confusion at INTC marketing while the lawyers figure it out. Combine the money AMD gets, the confusion INTC marketing is under, the general revulsion the consumer will feel at the publication of INTC's dirty tricks, the superiority of AMD's products and the loss of sales the INTC only OEMs are going to be feeling and you have a situation that could lead to explosive sales gains at AMD late next year.