SI
SI
discoversearch

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor.  We ask that you disable ad blocking while on Silicon Investor in the best interests of our community.  If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.
Politics : Formerly About Advanced Micro Devices -- Ignore unavailable to you. Want to Upgrade?


To: dumbmoney who wrote (63468)6/26/1999 1:20:00 AM
From: Yougang Xiao  Read Replies (1) | Respond to of 1572645
 
Fun reading: by Market_Shark from Yahoo.

This is the way I see it...Part 1
by: Market_Shark
44166 of 44177
You guys give it an ending. Have fun.

K6 got us in big financial trouble. Let me explain. It is a hell of a consumer
chip. But it infuriated the Gorilla. First they counter attacked with the celeron
(PII core-no cache). It was such a dog that we were even allowed to make
money 4th quarter of 98. Then the consumer reports came and INTEL did not
like them. AMD had over half of the consumer market. I can see manager in
charge, John Smith, being told “…crush the fly. I give you one quarter to fix it
or you are gone”. “No problem sir, consider it done” –the manager said. So
first thing, was to bump up the speed dramatically and add 128K of L2 cache
to Celeron.

Unluckily for us the K6 was never design to go past much 400mhz and sure
enough yields past 300 were close to non-existent. So, in a rush (2 months),
we tinkered with the process and got yields again. Amd's only fab was idle in
the meantime. Customers that were counting on shipments of 350-400 had to
be told to wait. Sanders makes the decision to sale VANTIS while they still
don't look desperate for money, and a decent price can be negotiated.

Meanwhile, a series of deep (up to 35%) price cuts kept pouring INTEL. At
times more than once a month.

We started smoking again, producing over 2 million chips per month in our
only fab. However, lost customers had no incentive to come back when they
could get celerons cheaper and cheaper. AMD kept its hopes up through
negotiation that orders would come back. They did. But customer after
customer demanded prices we couldn't meet. Inventory kept accumulating.
Well let's not jump the gun yet, we are introducing the K6-3 and that should
give us some leverage if we can price it just below PIII.

PIII, just introduced, was one speed higher than k6-3. By the time we
introduced K6-3 INTEL announced price cuts on PIII (shortest time after
introduction in history). Cuts so deep that PII processors became more
expensive than their newly introduced PIII counterparts. Of course the deepest
cuts were the K6-III matching speeds (even though the K6-III could not pose
much of a revenue loss threat to INTEL due to its large size and low
manufacturing quantities, it would have grabbed a fraction of a percent point of
PIII sales). K6-III found itself valued higher than a comparable INTEL chip
since introduction. Nobody (some were actually sold) were buying them.
Every time we tried to cut the prices on K6-IIIs INTEL would cut prices on
same speed PIIIs even deeper.

----continues in next post-----

We are in May and OEMs get celeron 400 for $80 and P-III 450 for $250.
AMD doesn't want to warn for fear of spoiling K7 introduction, get every
analyst to downgrade them and may be even loose customers due to a
heavily pessimistic painted business future. Besides, the quarter is still not
over and maybe a big sale could come and save the day. Let's not jump the
gun until we have to.

Sanders is very worried and accelerates the payment on VANTIS. It
happens. Gets word that K7s are yielding extremely well at 500-550-600
and that introduction will be on schedule. Dresden is on schedule. INTEL is
having problems with .18 coppermine. June is here. Inventory of K6 is
growing to millions. K6-III is not selling and Intel just announced more price
cuts.

Intel announces delay of coppermine, due to low yields at 600. INTEL gets
word that the shipping version of K7 is yielding extremely good at speeds
not yet achieved by PIII (at least out of a lab room). INTEL also gets word
that a consumer release-grade K7 system outperforms the fastest XEON by
a wide margin.

Management at AMD decides to make the quarter situation public at K7's
introduction to minimize shareholder damage and loss of business. K7 is
announced along with a pre-announcement of the poor quarterly financial
situation. K7 is out and is wearing the crown of Windows compatible
processors.

Manager John Smith at INTEL gets another call. “…Crash them again, your
job is on the line…”.
This time, he remains silent for a long second and replies: “With what?”

Sanders calls motherboard and chipset manufacturers. They are on schedule.
IBM reports on schedule. K6 production stops and all available fab space is
devoted to the manufacturing of Athlons and chipsets. Compaq reports on
schedule. Gateway reports on schedule. Athlons are starting to trickle and
finding home in the best-of-class soon-to-be-released computer and
workstation lines. The phone rings. “Hello, my name is John Smith,
remember me?. “Ah huh” replies Sanders. “Well, I just got a new job and I
am calling from Dell…”