Analyst Conference Summary

conference date: January 25, 2017 @ 2:00 PM Pacific Time
for quarter ending: December 31, 2016 (third fiscal quarter 2017, Q3)

Forward-looking statements

Overview: Growing, but not very fast. Good guidance for March quarter.

Basic data (GAAP):

Revenue was $586 million, up 1% sequentially from $579 million and up 3% from $566 million in the year-earlier quarter.

Net income was $142 million, down 8% sequentially from $164 million, but up 2% from $131 million year-earlier.

Diluted EPS (earnings per share) were $0.52, down 15% sequentially from $0.61, and up 6% from $0.49 year-earlier.

Guidance:

March 2017, fiscal Q4 revenue expected between $590 and $620 million. Gross margin 68% to 70%. Operating expense about $244 million. Net other expense $4 million. Tax rate 14%.

That leaves Xilinx on track to meet its annual fiscal 2017 guidance.

The revenue increase will be led by automotive applications.

Conference Highlights:

Moshe Gavrielov, CEO said "Xilinx delivered growth for the fifth consecutive quarter.  Sales from our 16nm Ultrascale+, 20nm Ultrascale, and 28nm Zynq products contributed to significant market expansion.  The growth from these products was driven by a broad base of markets such as data center, automotive, test & measurement, wired and wireless communications and space. . . In addition, our 16nm execution continues unabated.  We are currently shipping 12 unique products to nearly 300 customers and remain over a year ahead of the competition"

Revenues by end market: Communications and Data Center 44%; Industrial, Aerospace & Defense 41%; Broadcast, consumer and automotive 16%.

Communications led revenue growth.

Revenue by product type:

47% Advanced products: UltraScale, Virtex-7, Kintex™-7, Artix™-7, UltraScale+ (these are at 28 nm, 20 nm, and 16 nm)

53% Core products. So all the older, standard products.

"Xilinx introduced the Reconfigurable Acceleration Stack, with a goal to enable mainstream adoption of FPGAs in the data center. Target applications include machine learning, video transcoding and big data analytics. Designed for cloud scale applications, the stack provides the fastest path to realize up to 40x better compute efficiency with Xilinx FPGAs compared to x86 server CPUs and up to six times the compute efficiency over competitive FPGAs."

69.6% gross margin up sequentially from 69.2% driven by product mix.

Cash, equivalents and long-term investment balance was $3.25 billion. $1.0 billion long-term debt and $0.6 billion was current debt. Operating cash flow was $106 million. Depreciation $11 million. Capital expenditures $29 million. $214 million of stock was repurchased. Stock based compensation expense was $31 million. The dividend payment required $83 million.

Accounts receivable increased in the quarter, but is expected to normalize in the current quarter, resulting in improved cash flow.

Revenue by geography: North America 30%; Asia 43%; Europe19 %; Japan 8%.

Cost of revenues (GAAP) was $178 million, leaving gross profits of $407 million. Operating expense total was $244 million, consisting of: research and development $159 million; selling, general and administrative $84 million; and amortization $1 million. Leaving operating income of $163 million. Interest and other expense was $0.4 million, and the income tax provision was $20.7 million.

Tax rate was 13% in the quarter.

$0.33 dividend payable to stockholder of record on February 8, 2017 will be paid on February 23.

Q&A:

Broadcast, Consumer, Auto business down sequentially? Nothing anomalous in decline. It was more inventory positions of customers. We are seeing design wins, so no concerns, we expect a strong rebound in automotive.

Automotive has become the biggest end market, whereas it was infotainment two years ago.

How to think about op ex next fiscal year? We will give guidance at the next earnings release. We have been investing to exploit the technology leads we have, but are also committed to getting back to normal margins. We should be moving in that direction.

16 nm share gains, when might you hit $100 million run rate? It takes a long period of time from tech introduction to seeing huge revenue. We are still not at peak revenue for 28 nm products. We tend to look at 20 and 16 together as one node. They are doing well across all markets. For 16 none of the customers are in volume production yet. Automotive and datacenter will be two drivers, but right now datacenter is very early in its development, so the payoff is a few years down the road, say 2020. "I would be disappointed in a couple of years if 16 nm is not over $100 million in business."

In 2018 we expect to tape out and provide first silicon for 7 nm with TSMC. For the low-end portfolio we have expanded the 28 nm node.

Impact of Intel's 14 nm parts? We never underestimate our competition. But we are doing extremely well at 16 nm, with mature products and exceptional yield. We feel our competitive position is extremely strong at this node.

Growth trajectory for Zynq? Was for wireless, automotive, and industrial applications. It took longer than usual for designs to move into production. Wireless and automotive have ramped quicker. We have a mountain of design wins. Zynq is now 10% of overall revenue.

Datacenter revenue projections? We believe potentially hundreds of millions of dollars around 2020. We are just at the beginning, seeing design wins and interest, but not that much revenue yet, nor much growth in fiscal year 2018.

Datacenter acceleration was seen as a growth market by Nvidia years ago, and they invested in software to enable use of their GPUs. We need to do the same because the datacenter programmer base is not very FPGA savvy. We are working on providing help to customers, including software. We believe we can get to market relatively fast. We believe FPGAs can beat GPUs for many applications.

German auto maker switch from Zynq to Nvidia? Don't know who you are referring to, we have design wins with prior clients and new ones in automotive. It is early, too early to know which approach, CPU, GPU or FPGA, will work best.

OpenIcon Analyst Conference Summaries Main Page

Search

More Analyst Conference Pages: