<font color=blue>hey and pru projects upto 9mil cpus in q4
Someone is insane
07:46am EDT 10-Oct-00 Prudential Securities (H.MOSESMANN 650-320-1631) ALSC ALT
MICROPORCESSOR UPDATE - SUPPLY COULD OUTPACE DEMAND IN 4Q00/1Q01.
MICROPORCESSOR UPDATE - SUPPLY COULD OUTPACE DEMAND IN 4Q00/1Q01.
R E S E A R C H N O T E S October 10, 2000
Subject: Microprocessor Update
OPINION
========
Current: STRONG BUY
RISK: High
Analysts: Hans C. Mosesmann (650) 320-1631
Traci Tsuchiguchi (650) 320-1639
=============================================================================
- Given continued relative weakness in the PC end market, we believe the
prospects of a demand driven x86 processor glut are real in Q4/Q1.
- Supply, too many processors. We estimate Intel enters Q4 with an extra 2.5
million PIII class processors not sold in Q3. We are already cautious regarding
AMD's forecast of 7.2 million Athlon units in Q4. However, assuming they do
ship, given its manufacturing success, AMD may well produce over 9 million
Athlons in Q4 - Quite a few extra processors.
-ASPs could become a big problem. Weakness in PC demand plus increased
potential for processor oversupply in Q4/Q1 is a recipe for lower ASPs, which
means lower margins. While we believe Intel is better positioned longer term in
this battle, and with much of this dynamic already in both stocks, we would
rather take a wait and see position on both INTC and AMD in the near term.
- In fact, in the last week, we have seen rapid deterioration in the spot market
prices of AMD's Thunderbird processors, which declined between 3% and 23% from
the previous week. This compares to a more typical week over week decline
between 2-4%.
- Given the current environment, we continue to recommend stocks in the
near-term that have solid infrastructure communications exposure rather than
high PC-centric and wireless handset - centric markets. Our favorite near-term
top picks include: AXT, CREE, LSCC, MCHP, PSEM and XLNX. Longer term our
favorite remains TXN. Mid term plays that should recover nicely over the next
two quarters include: ATML, ALSC, EMKR, INSN, SIMG, and STM.
Microprocessor Update
PC demand clearly has not materialized as expected in 2H00 given the downside
pre-announcements by Intel, Dell, and Apple. Our industry checks confirm this
situation has not improved much. While Q4 should be up sequentially, our view
is that the first half of 2001 will likely lack catalysts for investors to focus
on in relation to both AMD and INTC. We believe it will be tough for AMD to
ship the 7.2 million Athlon units it has forecasted for Q4, up from the 3.6
million it likely shipped in Q3. While we believe AMD indeed has gained some
market share from Intel (Table 1.0), we simply believe 27% sequential unit
processor growth (without a major impact to ASPs) is not in the cards.
X86 Processor supply is now an issue. Given Intel's pre-announcement, we
believe Intel has an extra 2.5 million PIII/Celeron class processors in
inventory for Q4 that it had expected to ship in Q3. Furthermore, while AMD has
a shot of shipping the 7.2 million Athlons in Q4, their manufacturing execution
(Fab 25 and the new Dresden facility) has been so good that we would not be
surprised if AMD did not produce over 9 million Athlons in Q4. This would point
to a significant oversupply of high-end X86 processors in Q4, not to mention an
oversupply of them in Q1.
Average Selling Prices (ASPs) will be a critical issue. In addition to a
possible glut of high-end processors, which on its own is cause for concern, we
believe AMD's low end Athlon/Duron is a clear hit in the consumer market. We
expect this product could be the low-end processor product to beat in 2001.
However this success could turn out to be a mix problem for AMD given that more
sales of Duron relative to the higher end Athlon/Thunderbird may take a toll on
AMD's overall ASPs. Remember, the critical issue for AMD in 2001 is not
necessarily manufacturing execution (we believe they will execute), its
corporate market acceptance for the Athlon, which the company in general has
never been able to obtain. AMD needs to penetrate the corporate market in order
to achieve a more diversified end-market base which in turn leads to more
stable ASPs and we are not seeing evidence of much success on this front. On
another note, we expect AMD to enjoy very good upside potential in Flash
memories, however we would not view momentum in this area as a reason to either
buy or sell the stock in the short term.
Table 1.0
Processor Units (in millions)
1Q00(A) 2Q00(A) 3Q00(E) 4Q00(E) Year(E)
AMD K6 5.1 4.5 3.4 1.9 14.9
AMD K7 1.2 1.8 3.6 7.0 13.6
AMD (total) 6.3 6.3 7.0 8.9 28.5
Intel (total)31.0 32.5 32.5 35.2 131.2
Total 37.3 38.8 39.5 44.1 159.7
AMD Share 17% 16% 18% 20% 17.8%
Intel Share 83% 84% 82% 80% 82.2%
AMD QoQ Growth 0% 11% 27%
Intel QoQ Growth 5% 0% 8%
Total QoQ Growth 4% 2% 12%
AMD ASP $ 88 $ 93 $99 $104 $97
Intel ASP $ 190 $ 190 $190 $190 $190
Source: Company reports and Prudential Securities. |
