Here's my transcript from AMCC's conference call: (all errors mine) Incidentally, every analyst on the call congratulated the company on the quarter and some even on positive guidance.
>>>>>>>>>>>>>
AMCC third quarter CC, Jan. 16, 2001
[See press release for financials.]
Book to bill: 1.25; $179.4M in bookings, $186.2M if MMC were included for entire quarter.
$215.6M in backlog as of Dec. 31. $195.4 shippable within 6 mos.
Gross margins 76.3%, vs. 74.8 last quarter.
21.6% R&D, up b/c of MMC
13.9% SG&A, flat sequentially
Operating expenses 38.6% vs. 41.8% seq., and 37.6 y/y
$14.8M in interest income
$13.5 M in future Qs b/c of interest rate declines
$56.B in assets
$1.13B cash and short term
Inventory turns 6.0 vs. 8.8
DSOs 55 days vs 45 days --- 3 days were cosmetic b/c of MMC acquisition, the rest b/c of IBM supply constraints
Share count --- 323M vs. 306M, estimate 6M per quarter increase
* * *
Strength in OC192 framers
50% of communications revs was from products introduced in 2000 or later.
OC192 ramped from $14.0 M to 26.9M y/y --- 79%
60% of OC192 was to optical module mfgrs, 40% to OEMs.
112% rev. increase sequentially.
Supply constraints have begun to clear.
Have 21 customers over $1M up from 5 a year ago. New customers include ONI, NEC, Sumitomo, Unisphere, Astral Point, Cyras, Network Systems (?) and Yahowah (? --- Chinese co., someone correct my spelling, please).
NT at 18% is largest customer, this excludes OC-192 products that go to module customers and eventually end up at NT. 68% of NT reven is from standard products vs 16% a year ago.
Second largest customer is CSCO at 10%.
Rev strength in standard products. (This appears to be a different part of co., I don’t follow so need help here.)
Bookings BtoB 1.25 in Dec., communications were 1.33, and OC192 of $51.6M were 1.9 BtoB
Diversity of customers, 25 over $1M versus 7 a year ago.
First OC768 has been ordered, to be shipped later this year. Significant revenue a couple years away.
* * *
Operations:
Important optical customers coming on. CSCO improved growth going forward. We managed CSCO business well and have entered Mar Q with no excess in system. (MMC)
NT up 30% sequentially. Shasta came back nicely.
MMC --- 10 design wins, including 3 at CSCO. In second half 2000 we had 25 wins, with 11 in optical. We now have 34 total in optics, a “highly strategic area.”
Merger going well. Have targeted metro edge. OC-48 and OC-192 roadmaps set out. Beginning design on architectural interface. Customer response is excellent.
Optical Market perspective:
We see very robust growth. There are four factors:
1. Majority of revenues comes from fastest growing part of market
2. Mix of optical shipped is at highest speeds
3. Add dollar growth at all levels
4. Take market share
To expand:
1. Long haul transport and core routers --- 2/3 comes from core. RHK states optical core grew 90% in 2000 vs. 37% the year before.
2. Mix to higher speeds: in 1999 90% was OC-3 or OC-12. Now more OC-48 and OC-192.
3. Adding dollar content at each speed. Last year added $73 per port.
4. Optical was up 90% in 2000. Went from 13% of revenues to 18% between 1999 to 2000.
We weighted each factor as to how it impacted AMCC:
35% of growth impacted by number 1, 18% by number 2, 26% by number 3, and 21% by number 4. Because of this, AMCC will grow faster than market.
See continued strong market in optical in 2001. Bullish on outlook for quarter.
* * *
We faced several challenges this past Q and powered through them. Peripheral softness at CSCO was offset by other customers. We also overcame decline in non-communications business.
Gross margins 76% for combined entity. 67% of quarter was outsourced vs 55%.
We integrated several new products. OC-192 “Hudson” --- adopted by major WAN manufacturers.
Outlook: Inventory, capex, and economy --- we would be foolish and arrogant if we were not concerned. Demand has remained strong. No indications of a slowdown in our business. We had a supply issue at IBM that is easing up. Turns business historically has less visibility. We have a good track record, but it’s possible we could be surprised. We see sequential growth in the upper teens to 20% in Q4. This is from a basis of $150M in Q3 (calculated as if MMC had been part of co. all quarter). Q1 at traditional 10 to 15%.
Put gross margins at 75% just to be safe. R&D at 22% and SG&A at 14%.
Macro economics --- we haven’t seen effects at AMCC.
Q&A
Q: Why didn’t you ship more OC-192s?
A: We could have but IBM was at or above capacity. We have some delinquency. SiGe had pinpointed constraints but now running olk. Wafers had spot problems.
Q: Are substrates being shipped?
A: We talk to them every day. Seeing capacity freeing up.
Q: Any success bringing up numbers in IBM product line?
A: Looking at second sourcing. IBM is also second-sourcing.
Q: Any in OC-48?
A: Got caught up late in Q. Some impact on framer business. This Q will be tight
Q: Guidance re: MMC and AMCC?
A: Looking at total business. Communications will carry the day. AMCC growing faster than MMC.
Q: OC-192 framer designs --- what market share?
A: Just getting started. No models yet.
Q: Over 50%?
A: Have most OC-192 framer market. We’re only commercial company doing this now. Others coming along. (commercial vs ASIC)
Q: Customer inventories? What is impact of shift to foundries?
A: Have scrubbed forecasts. Looked at subcontractors and asked if they have stock-piled. They said not in our line. We challenged every line entry, line by line. We think there is upside to bookings. Less downside than usual b/c of rigor of scrubbing.
Q: Foundry impact?
A: Addition of OC-192 --- think gross margins won’t deteriorate. It will be a long time before competitors affect our products.
Q: OC-192 revenues?
A: OEM was low because early entrants are module makers. Later OEMs will ramp. 25% of OC-192 was framers.
Q: Tier One accts?
A: We don’t break it out. High percentage of big guys and smart start-ups. Big guys are adopting. It’s a mis.
Q: NT --- larger mix --- how is base business? New products?
A: Large % was AMCC kicking off in addition to MMC. Yes, was up.
Q: SiGe?
A: 24% SiGe, 42% CMOS, rest bi-polar.
Q: Distribution-end inventories?
A: No basis change.
Q: What percentage of sales was constrained?
A: Framers more than SiGe. Framer digital was constrained. We could have done 30%seq, but not 40%.
Q: Lead time risk?
A: Don’t think there is stock-piling. We juggle to keep customers happy.
Q: Do you see fab utilization increasing?
A: A lot of growth is from outside. Internal will be flat b/c more products are in outsourced area. Internal fab is operating at peak levels.
Q: What about optical, Hudson?
A: A lot more in OC-192 in both framer and PHY. Newer switching fabric. From small to terabit. A lot more in optical end. Broadening at OC-192. More into optical access. On Hudson we’re enabling the technology. We have forward error correction. This is a big highlight in this area.
Q: Uniline fabric?
A: Nice design wins. Customers like fabric architecture.
Q: Uniline plus MMC?
A: Putting architecture interface for uni fabric. Also lower density fabric --- broader range. We offer compelling solution. Customers want integrated solutions.
Q: Competitive landscape? New entrants?
A: A lot are salivating over our markets. Re respect PMCSierra. We’ll be watching each other. Also VTSS, BRCM, CNST, INTC --- there are a lot out there. But we’re confident.
Q: What keeps you awake at night?
A: You guys! Wall Street! Rumblings that could become self-fulfilling. I can’t control it. Haven’t seen it. Keep waiting for it to bite me.
Q: What about downstream? Turns?
A: Working on framers at OC-192. Want seamless connection from network processor to framer. Higher layer will pull lower.
A: Turns --- 25 to 30%. Some benefit from IBM supply easing up.
Q: 60% in optical module business. You are early to this space. Does it change anything?
A: It is a little difficult. Less complicated system. They spin modules more rapidly. Very little slop in design schedules. Tremendous leverage in this business. Going forward, they’re growing and adding capabilities. Very strategic. OEMs look to module guys as leaders. OEMs will ramp as % of total. |
