Currently, AMD (AMD & Fujitsu) has surpassed Intel as the number 1 producer of flash. This
quarter, we estimate AMD will have flash sales in excess of $260 million, and
will be on a $1 billion run rate for the next 2-3 years.
Courtesy of Albert:
07:02am EST 15-Nov-99 PaineWebber (Lazlo, John J. 415-576-2980) AMD
AMD: Analyst Meeting
Semiconductors PaineWebber
John Lazlo RESEARCH NOTE
415-576-2980/lazloj@painewebber.com
Ted Parmigiani, Associate Analyst (415-576-2974) November 15, 1999
Advanced Micro Devices Rating: Neutral
(AMD-$26.56)
AMD: Analyst Meeting
KEY POINTS
* Revised Q4 top-line estimate to $913 million, up from $700 million, to
reflect better-than-expected outlook. Company should be on track to ship
roughly 1 million K7-Athlon units as compared to our previous estimate of 500
thousand, which was based on a then-anticipated shortage of chipsets and
motherboards. The Communications business is expected to post 20% revenue
growth in the current quarter versus our previous estimate of 15%. We hiked
our flash memory revenues from $220 to $260 million.
* We raised our current quarter EPS estimate of a $0.54 loss to a slight profit
of $0.02. For calendar year 2000, we believe the company could earn $1.00
versus our previous estimate of $0.40. This reflects our estimate for a $0.05
better than breakeven first half 2000, followed by roughly $0.95 EPS for the
second half of the year.
* We reiterate our Neutral rating but raise our price target from $16 to $25 as
the company progresses towards short-term breakeven and longer term
profitability.
Key Data Quarterly Earnings Per Share (fiscal year ends
December)
52-Wk Range $32-15 1998A 1999E Prev 2000E Prev
Eq.Mkt.Cap.(MM) $3,931 1Q ($0.39) ($0.81)A ($0.03) (0.31)
Sh.Out.(MM) 148 2Q (0.45) (1.10)A 0.08 (0.08)
Float 70% 3Q 0.01 (0.72)A 0.36 0.27
Inst.Hldgs. 44.2% 4Q 0.15 0.02E (0.54) 0.59 0.52
Av.Dly.Vol.(K) 2,552 Year ($0.68) ($2.59) ($3.16) $1.00 $0.40
Curr. Div./Yield None/NA FC Cons.: NA ($3.02) $0.56
Sec.Grwth.Rate 15% Revs.(MM): $2,542 $2,802 $4,698
12-mo. Tgt Price $25.00 P/E: NM NM 26.6x
12-mo. Ret. Pot'l (5.9%)
Convertible? No
1999 ANALYST MEETING
We attended the company's annual analyst meeting last Thursday. The meeting was
upbeat as the company's update on how business is tracking in the current
quarter was much better than expected. Besides this update, the company
discussed their flash memory and communications products business, their MPU
business and roadmap, and capacity expansion for the short and long term in all
product segments.
The company's flash memory products are sold out for the current quarter. Q4
will be the seventh quarter of sequentially higher unit volume and revenue in
this business segment. This is representative of an industry-wide shortage of
flash primarily due to burgeoning end market demand for cellular handsets and a
limited number of producers. Other end markets for the company's flash products
include set-top boxes and handheld games and devices. Going forward, we believe
the company has planned to more adequately meet oncoming demand from these high
growth end markets via its joint venture, FASL, with Fujitsu. Currently, AMD
(AMD & Fujitsu) has surpassed Intel as the number 1 producer of flash. This
quarter, we estimate AMD will have flash sales in excess of $260 million, and
will be on a $1 billion run rate for the next 2-3 years.
The company's communications business is also experiencing positive sequential
growth, and guidance was given for a 20% gain over the previous quarter. We
revised our estimate of an 8% gain or $95 million to the guided revenue of $120
million. We believe this segement should generate annual revenue in excess of
$400 million. However, as was recently announced in the company's 3Q conference
call, AMD has made plans to sell this unit before the second half of 2000.
AMD's MPU roadmap was discussed, with an emphasis on planned die shrinks,
increasing clock speeds and next year's competitive strategy focused on gaining
market share at the higher end of the MPU market (workstations and servers).
Currently, the majority of AMD's MPUs are made using .25-micron process
technology. Within this quarter, the company will begin production of its .18-
micron K75-Athlon and its K6-2+ and K6-3+ MPUs which will also be made using
.18-micron process technology. Starting in mid-2000 and using the company's Fab
30 in Dresden, K-7 Athlons will be made using .18-micron with copper process
technology. This will be possible with a previously announced joint alliance
with Motorola, which will supply AMD with the HiP6L (6 layer metal dual copper
inlaid technology) technology, enabling clock speeds in excess of 1 GHz.
Subsequent die shrinks and clock speed enhancements will be made possible by
the HiP7L (.13-micron) and the HiP8L (advanced work in transistors, lithography
and interconnect) to be in production some time in 2001. Fab 30 is expected to
ramp to full volume in 2001. Fab 25 in Austin should be fully converted to .18-
micron in 2000 for the K-6 family of MPUs and will produce a mix of K-6, flash
and general purpose logic (sub .25-micron).
While the AMD MPU roadmap is competitive with that of Intel's, we believe the
greatest challenges that AMD will face going forward are in the areas of
execution and marketing. Fabricating MPUs using copper is widely believed to be
more complicated than the current alternative and could delay volume shipments.
This was evident in Motorola's delay of the Apple G4 processor which is also
made using copper interconnects.
From a marketing standpoint, AMD has already proven to be a viable competitor
of Intel in the value PC segment as seen in Intel's aggressive pricing strategy
of Celeron to compete with AMD's K6 family. But in the high end desktop,
mobile, workstation and server MPU segments, Intel is "inside" today and will
aggressively go after more of this higher end market going forward. That said,
AMD will have to overcome the overwhelming acceptance of Intel Architecture by
corporate IT executives who base there future MPU platform decisions on best
overall performance with the least total cost of ownership. Our sense is that
Intel is a better bet than AMD as the company which will continue to gain
market share in the high end of the MPU market.
OUTLOOK FOR Q4 AND BEYOND
Strength is being seen across all product segments: flash is sold out,
communications products should post a 20% sequential revenue increase, and the
company is confident it will ship 1 million or more K7-Athlons due to better
than expected chipset and motherboard availability as well as successful
execution of its next generation MPU production. The flash memory market is
experiencing strong demand with a limited number of key suppliers (e.g. Intel,
AMD, and Atmel). In short, we have revised our Q4 top-line estimate to $913
million, up from $700 million, to reflect better-than-expected results in all
product segments.
We raised our current quarter EPS estimate of a $0.54 loss to a slight profit
of $0.02. For 2000, strength is expected to come from flash memory products,
and the probable sale of the communications business could generate a one-time
gain (not factored into our model) of at least 1-2x annual segment revenues.
The success of the company's core MPU business will greatly depend on execution
of roadmap and marketing strategy. Our estimates assume moderate success with
respect to these issues, and we believe the company could generate EPS of $1.00
for 2000 versus our previous estimate of $0.40. This reflects a $0.05 EPS in
the first half of 2000, followed by roughly $0.95 EPS for the second half of
the year.
INVESTMENT SUMMARY
We reiterate our Neutral rating on AMD and raise our price target to $25 from
$16, which suggests a 25x P/E times our calendar 2000 EPS projection of $1.00.
We believe that the shares will trade in this range until there is evidence
that AMD has demonstrated its ability to return to profitability. Moreover,
investors must gain greater confidence in management's ability to execute its
K6 strategy. The design of the K6 and its progenitors Sharptooth and K7 are
excellent products with performance characteristics that rival the comparable
Intel parts. However, AMD has not been able to reliably deliver the parts to
the same extent that Intel has executed. Intel is also shaping the direction of
the new computer industry and its R&D spending is roughly the same size as
AMD's revenues. While we believe that that there is room for at least two
players in the MPU space, Intel appears to be the more solid investment
vehicle. While Intel was late to enter the sub-$1,000 PC market with its
initial version of Celeron, it has regained some of its lost share with its
second generation Celeron product (Mendocino). As a result, Intel may continue
to price the Celeron aggressively in an attempt to gain further market share in
this space. This strategy could make it difficult for AMD to sustain
profitability on a long term basis.
RISKS
Risks include the continuing production ramp-up of the next generation K7
(Athlon) chip, the short-term availability of chipsets and motherboards, the
demand for microprocessors, competitive pricing pressures from Intel, and a
continued recovery in demand in non-microprocessor areas such as communications
and flash memory products. |
