conference date: July 24, 2014 @ 1:45 PM Pacific Time
for quarter ending: June 30, 2014 (Q2, second quarter 2014)
Overview: Strong growth quarter, in contrast to competitor Xilinx.
Basic data (GAAP):
Revenue was $491.5 million, up 7% sequentially from $461.1 million and 17% from $421.8 million in the year-earlier quarter.
Net income was $127.0 million, up 9% sequentially from $116.5 million, and up 25% from $101.5 million year-earlier.
Diluted EPS (earnings per share) was $0.41, up 11 % sequentially from $0.37 and 32% from $0.31 year-earlier.
Q3 sales expected to be flat to -2% to 2% growth. Gross margin 66.5% to 67.5%. R&D expense $114 to $116 million; SG&A $78 to $80 million. Other expense $4 illion Tax rate 11% to 12%.
By vertical: telecom and wireless revenue flat; industrial, military and automotive flat; network and computer up; other down.
Almost all of the wireless growth in the quarter was outside China. LTE in China is likely to decline in Q3.
Server market is becoming a bigger opportunity for Altera. Microsoft announced plans to use Altera chips to accelerate Bing searches.
Continues to work towards products based on Intel's 14nm Tri-gate technology, with test chips already back. Believes can replace more ASICs with this technology.
Cost of sales was $162.4 million, leaving a gross margin of $329.1 million. Operating expenses of $182.6 million consisted of: R&D $101.1 million; selling, general and administrative $79.0 million; and amortization of $2.5 million. Leaving an operating margin of $146.6 million. Interest and other expense was $3.0 million. Income taxes $16.5 million.
Revenue by geography: Americas 16%; Asia 42%; EMEA 27%; Japan 14%.
Revenue by vertical market: Telecom and wireless 46%; industrial and military 21%; networking & computer 15%; other 18%.
New products contributed 53% of revenue; mainstream 21%; mature and other 26%.
84% of revenue was for FPGAs, 8% for CPLDs and 8% for other products.
Cash, equivalents, and long-term investments ended at $4.5 billion. Long term debt was $1.5 billion. Cash flow from operations was $171.0 million. $9.6 million was used for capital expenditures.
Believes can become the market leader with its next generation products. 14 nm Stratix family design on schedule; believes will capture the high end of the FPGA market with this. Believes having 20 nm design software available is an edge over the competition.
Magnitude of Q3 wireless correction? Just down in Q3 and Q4, after being up in Q1 and Q2. We are seeing buildouts outside of China. Because of a shortage in other parts (analog power amplifiers) customers built a backlog of Altera chips, particularly China Mobile. China telecom LTE should pick up at some point.
Computer opportunity timing? We believe it will be up sharply in calendar Q3, but it will be choppy because of the way datacenters are built. But we would expect it to be up y/y.
China LTE, is it over? It is very early in the LTE build. In the U.S. even though Verizon and AT&T started build outs 2 years ago and then paused, they are having to do a second round of build now. In China it has been just China Mobile, and only one phase of that. They may halt for Q4 but then restart some time in 2015. We believe wireless has a lot of potential going forward.
28 nm business, impact on margins? It is the same as for other node ramps. Margins are depressed at first, then improve as the product matures. But wireless has lower gross margins than other markets, and this year a lot of growth has been from wireless. We are at the low end of the margin range today.
We had about 35% market share last quarter. We would expect to get to 40% this quarter and expect to gain market share for the full year in programmable logic.
Change in momentum, you vs. Xilinx? Our product portfolio is fine at the low end and advantaged in the high end. In the midrange we have a slower product line, and the competition's product matched market requirements better than ours. Midrange came out of the gate fastest, for radios. The higher end is in test and measurement, base stations, and telecommunications, which require longer to ramp, and is happening now and helping us take market share.
Sampling timeline for 14 nm? Q1 for first production part. Software is already going out to early access customer. New product gave one customer 2x performance over competing products.
PLA devices have long tails at each node. Older products continue to chip. China LTE is a combination of 28 nm, 40 nm, and even 60 nm. We believe we are number one in wireless, so we have received a bump from wireless segment growth.
We are shipping into the announced 500,000 China Mobile LTE base stations planned for this year. We don't know their plans for 2015. The Chinese competitors were given licenses later, and limited compared to China Mobile.
We see FPGAs becoming more important in the server market partly as a result of power costs being the primary operating costs. Also they can do math 10 times as fast. But we had to solve the programming flow. That is why Altera developed a compiler to go from computer languages to FPGA program code for math algorithms including search, image compression, and encryption. Anything where the math is critical. These tend to be the high-end, most expensive FPGAs. But there is competition from GPUs as well.
We are changing architectures as well as moving to true 14 nm, which allows us to speed up the fabric and potentially replace more of the ASIC market. But we are not seeing increased ASIC activity yet.
Design activity in wireless market? It is very aggressive, particularly in radios, which tend to get redesigned constantly. The ARM microprocessors embedded in the chips are becoming more important in wireless, and we are the only supplier for that.
Analyst Conference Summaries Main Page