Chase H&Q Conference
SDLI, Don Scifres, CEO
Fiber optics business grew 163% y/y, 35% sequentially
80 PhDs and 1500 employees, largest technology group in the industry. Large patent portfolio, large customer base --- commanding 67% sequential growth to bottom line.
Expanding facilities at all sites: 50% more in San Jose, 2X more in B.C., Canada, 100% more in MA over next 3 months, 50% more in UK, 3X or more at Veritech over next 6 mos.
Undersea business was less than 1% of revenues 15 mos ago, as of last Q it is 30%.
Gross margins increased 7 pts in one year. Grew operating net from 16% to over 30% in twelve mos.
We get out ahead of market trends. We're integrating more value into products. Now 50% of revs is in modules.
Looking at all optical networks.
980 nm pumps
EDFAs
Raman laser modules --- first qualifications now include 13 customers
Lithium Niobate drivers and modules
TxRx (transmitters, receivers) --- Veritech recognized for undersea, yet new customers coming from terrestrial.
Network Monitoring ---Queensgate --- look at wave spacings in each fiber
Looking to expand products further --- you'll hear more over next several months.
Raman 1.7 W --- record power level and only one that's telco qualified.
$1.9 billion market by 2004 from $10 million in 1999.
Pumps --- we are obsoleting ourselves. New chip and new package for next generation --- 500 mW, 400 mW, 300 mW.
Modulators and attenuators grew 157% y/y. Ramping manufacturing.
Receivers --- sell as group or as individual. 3X floor space needed, 130,000 sq. ft. up from 18,000 sq. ft. today.
Network Monitoring --- now coming into play. $10m market in 1999, estimated to be $1 billion going forward.
1Q 83 % of revenues from fiber optics and 17% from printing and industrial lasers. Fiber optics will continue to grow.
We gain early access to customers. Adding value at module level. More customers are outsourcing.
Demand is exceptionally high:
* increasing facility space
* improving yields
* running 24/7
* more automation
We're partnering with leaders --- target high growth areas --- Alcatel, Corning, NT, Lucent, JDSU, Agilent, Antec, Tyco, Pirelli, Ciena, etc.
Seeking to expand
Financials: had great quarter and we expect more to come.
Goal is 30% operating income.
We are targetted on next generation of products.
Q: Further improvement in yields?
A: Depends on product line. We see possible 2X improvement in 980nm.
Q: Time table for adoption of Raman?
A: Undersea in production now. Terrestrial --- 2001 for 40 gig. But terrestrial will be big --- deployments are 1 year away.
Q: Price?
A: $50,000 to 60,000 each
Q: What about packaging?
A: There are 15 different technologies necessary to make module-level packages --- packaging needs sub-micron alignment and must stay aligned. 8 or 9 components fit into one pump module --- must improve yields. New package will differentiate product line. Strong relative to competition.
Q: Gratings?
A: We bought C-Star Optics in 1995. We focus on next generation of products. At that time gratings weren't so much an advantage. With EDFA, you needed stabilization. It wasn't so important with 4-channels. When you have 80 or 160, you have to have flat gain over many channels. Stabilization is critical. It took 3 or 4 years to be a differentiator. Now 70 - 80% of 980 nm pump market is SDL's --- and it's the fastest growing part of the market. Satellite business was toughest for reliability specifications. Submarine is the second toughest. We have 50% of 980 nm submarine.
Q: What about JDSU as customer?
A: Demand is strong. So no change. In fact pump modules grew 45%. We're negotiating contract to 2001. As JDSU becomes larger, some of NT and Lasertron's customers are looking at SDLI --- as JDS becomes more integrated.
Q: Market share?
A: We have 50% --- pump modules 40%, undersea 65-70%, Raman over 50%. Our target is to be number one in every market we address. We just acquired Queensgate, the best network monitoring system available today, monitors both C and L Bands. No one else can go to 25 GhZ, 256 channels. Getting signed into many high data rate systems.
Q: Competition, specifically Digital Lightwave?
A: Queensgate uses Digital Lightwave. We are taking market from electronic monitoring. By tuning spacing of etalon, you determine wave length; tunable wave length filters will be base of new products.
Q: Metro market?
A: We're modifying some products for Metro market. Raman going into some. Metro will use basically same type of equipment. We have 980 nm pumps for $1000 for this market.
Q: Rollout of 300 mW 980?
A: In hands of customers now. Product in 3Q. Will phase in over time. By end of year hope to have most product --- 200 mW will be low end, 300 mW will be high end. 1 watt chip in 2001. Chip operating 300 mW doesn't have cooler high enough yet. 400 mW being qualified for underwater now. (Cooling not an issue.)
Q: All-optical networks --- what's strategy?
A: We have 1) Queensgate 2) Veritech --- low noise receiver for submarine is also great for metrol, 3) Raman --- allows transparent fibers (demonstrated with ATT at OFC) 4) all-optical doubles need for amplifiers. Submarine are every 100 km, terrestrial must be 50 km -- or else use Raman. 980 nm is lowest noise pump --- so 2X needed in metro market.
Q: Corning?
A: Business going well. No one wants sole-source because of huge demand. NT buys 70% of product outside. Right now we're primary source in high power --- new package will differentiate us from competition.
Q: Free-space channels?
A: Pump for free-space channels --- EDFA could go into terabeam systems.
<<<<
I've noticed many analysts commenting on valuations and there seems to be an over-all concern that the dot.com phenomena is causing all stocks to be re-evaluated.
From the semiconductor panel:
Q: How will we know if the cycle turns?
A: Key variable is demand. If demand is okay, inventories won't build up. Look at companies close to end-user.
Q: Will fed's raising rates put pressure on demand? When will capacity catch up with demand?
A: Don't see it this year or next. Pretty tight for next 2 years. Won't see huge capacity coming on board b/c of long lead times, i.e., steppers take 12 mos.Optical components are a different proposition. Demand is phenomenally high.
Q: PC as driver?
A: This cycle hey're not as much a driver. Last year 25%. Base of applications is expanding, like cell phones. Still flash constraints. Second half will be good.
Q: Outsourcing trend?
A: System vendors are outsourcing to component suppliers. Your'e selling an architecture. Abrizzio router from PMCS --- they put it together to understand it. If you're optical and out outsource everything --- like Sycamore --- you're really an electronics company. Just look at architecture. Active components follow cycles similar to semis; passives components --- this doesn't apply. Lends to outsourcing, ie.e, ETEK in China.
Q: What about Intel?
A: They're 15% in telecom. Trying to increase. They were late to realize PC was not the big driver going forward.; They have a lot of challenges. PMCS and BRCM are system companies with chip expertise. INTC not on the DSP "map." Why would Nokia, for example, buy from Intel? They need a software platfor for carrying forward. Hard to switch one you're designed in. Hand-helds are driven by micro-controllers, not micro-processors.
Again, watch equipment company's growth.
Q: Who are leaders?
A: ASM, AMAT, Altera, Xylinx, Varian Semi.
Q: What about valuations?
A: No way to justify these valuations --- eventually things come back to those that make sense. Look at best companies in highest growth areas. There'll be further rotation into semis --- market won't sell off on valuations alone. We could see multiple expansions in this flight to quality.
Small cap: Xico (XCOR)--- (got strong recommend)
Message 13623267
Varian --- [VSEA] another that's undervalued. Nobody has growth without them.
Micrel is another that's undervalued. We'd be traders here (speaking of all semis re: valuations).
What about RMBS? One analyst: They'll make a lot ofmoney. Another analyst: There are cheaper, easier ways to play the cycle.
>>>
That's it for now.
Tomorrow JDSU presents at 8:30.
Pat |
