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.
Technology Stocks : Advanced Micro Devices - Moderated (AMD) -- Ignore unavailable to you. Want to Upgrade?


To: andreas_wonisch who wrote (18612)11/10/2000 3:34:51 PM
From: MaverickRead Replies (1) | Respond to of 275872
 
ML:AMD will sell out Athlon & Duron in 4Q00
Excerpts from ML Research 11/10/00

Investment Highlights:
• We attended AMD’s Analyst meeting
yesterday. Management presented on both of
its largest business segments, flash memory
and microprocessors. Most of the attention,
however, was centered around AMD’s flash
business.
• AMD staunchly defended its belief in strong
end market demand for flash and suggested
that its business strength will continue for
several years. We believe that increasing
density and tight supply will continue to
increase ASPs through 2001.
• While we believe that AMD continues to enjoy
good performance in its microprocessor family
- shipping 28 million units in 2000 and selling
out of both Athlon and Duron processors this
quarter - the company has confirmed that
overall PC demand this quarter has been
disappointing.
• We believe that the company is on track to
deliver $4.8 billion in revenues for 2000. Our
estimate are unchanged, $0.74 for 4Q, $2.56
for 2000. Our recommendation is an
Accumulate/Buy.

We attended AMD’s Analyst meeting yesterday. We
believe that the company is on track to deliver $4.8 billion
in revenues for 2000. Management presented on both of its
largest business segments, flash memory and
microprocessors. Most of the attention, however, was
centered around AMD’s flash business.
Our estimate are unchanged, $0.74 for 4Q, $2.56 for 2000.
Our recommendation is an Accumulate/Buy.
 Reassurance on the flash business
With much attention focused on the future prospects of the
phenomenally growing flash industry, AMD staunchly
defended its belief in strong end market demand for flash and
suggested that its business strength will continue for several
years. We believe that increasing density and tight supply will
continue to increase ASPs through 2001. AMD’s
management believes that cellular handset shipments should
number 575 million next year and that the supply/demand
imbalance will continue into 2002. The company did mention
a few instances where customers pushed out flash orders.
However, we do not believe that these examples were
indicative of general market demand and note that AMD was
able to resell the products to other customers.
Building sufficient capacity to meet demand continues to
be the main focus for the company. The company
continues to add significant capacity with its Fujitsu JV.
JV2 fully ramped flash production during this year, while
its Iwate and Gresham fabs continue to ramp. The
company’s JV3 facility should bring additional capacity in
2001. AMD expects to increase flash capacity by 77% and
ship 350mm units next year. We expect demand for flash
will grow over 100% YoY.
 Fourth quarter lags; the return to normalcy in
2001 for PCs?
While we believe that AMD continues to enjoy good
performance in its microprocessor family - shipping 28
million units in 2000 and selling out of both Athlon and
Duron processors this quarter - the company has confirmed
that overall PC demand this quarter has been disappointing.
We have heard that excess inventory levels in the third
quarter have spilled into fourth quarter throughout industry.
The fourth quarter’s typically strong Christmas sales have
yet to materialize. After this tough second half for the PC
industry, AMD believes that next year should follow more
“normal” seasonality, with PC units shipments growing in
the mid teens YoY and microprocessor unit shipments
growing in the high teens YoY.
The company’s MPU growth prospects in 2001 are helped
out by several factors in its roadmap. Not only will AMD
have its new line-up of processors in place, but also it will
have the core logic platforms for its processors. DDR
chipsets for the high-end are already sampling and will
begin to ship in volume early next year, while the low-end
UMA chipsets should be available by then, as well.
We believe the company’s assertions that its MPU ASPs –
currently $90 - should remain around the same levels next
year. We expect the sales mix to shift to more high-end
processors, which should firm up ASPs.
 Microprocessor Roadmap into 2002
AMD revealed more details about its microprocessor
product and technology roadmap into 2002. In 1H01, we
expect to see upgraded versions of its Athlon processors.
Palamino, the high-end processor, will reach 1.5 GHz by
2Q01 and will be manufactured on 0.13 micron process by
4Q01. The low-end version, Morgan, will trail the
Palamino by a quarter. The release of mobile versions of
these processors will fill an existing gap in AMD’s line up
in the mobile space. With the Palamino and its Hammer
family, AMD will have the ability to enter the low-end
volume server/ workstation market. However, we believe
that AMD’s penetration into the market will take time, as
this is an entirely new market for the company.
Later in 2001, AMD expects to sample its Hammer family
and ship in volume by 1Q02. The Clawhammer will target
the mainstream market of 1-2 way servers, workstations,
performance desktop and mobile products, while the
Sledgehammer line will target the higher-end market
including 4-8 way servers. We believe that AMD will have
the capacity to bring on these new products as its Dresden
fab will ramp to full production next year with SOI and
0.13 micron process capabilities.



To: andreas_wonisch who wrote (18612)11/10/2000 3:37:55 PM
From: rsi_boyRead Replies (1) | Respond to of 275872
 
I don't think we'll see 5 12 until we go to 0.13
improving cash bandwidth or improving branch prediction would probably achieve greater performance benefit than increased cache sizes and with way less heat/diesize/manufacturability penalty

</speculation>

t.