The past two days Xilinx and Altera have each reported quarterly results for their June quarter. The market reacted much more positively to Altera’s results sending the company’s shares up 12% yesterday. This is largely due to Xilinx long-term strategy still being fairly opaque. Overall both company’s results were quite good. Altera had a great quarter and in many respects appears better positioned. However, Xilinx may be building the ground work for a game changing announcement later this year.
Earlier this year I outlined how digital products are increasingly migrating to Xilinx and Altera’s FPGAs due to the strong platforms offered by these companies; enabling convergence in semiconductor design. Let me provide an update based on the June quarter results from these two highly profitable companies who control over 80% of the FPGA market.
Altera won this quarters battle
Historically Altera has had better margins and a much higher return on equity. A week ago Zack’s semiconductor analyst Abdul Saleh put out a timely buy recommendation on Altera referencing Altera’s:
- HardCopy structured ASICs, the leading structured ASIC products
- Partnership with Taiwan Semiconductor (TSM) which is allowing them to leapfrog Xilinx to 40nm
- And of course beating the drum of FPGA’s replacing ASICs
Along with these points I think Altera’s growth will be driven through its partnership with ARM (for software processors) and its low power leadership. The partnership with TSM is a big deal. Altera and Xilinx are both fabless which means they outsource the actual manufacturing of their chips. TSM controls approximately half of the semiconductor manufacturing market. Intel has historically had the premier manufacturing facilities. TSM’s foundries are starting to have nearly equivalent capability demonstrated by their ability to produce 40nm FPGA’s later in the year.
Novice note: The smaller process node (for now fewest nanometers) a company can use to manufacture their chips the more powerful the chip can be.
Altera’s Q2 2008 Results (released July 15, 2008)
Financial Highlights
“The combination of solid top line growth, continued gross margin improvement, and lower than planned operating expense created significant operating leverage…We are on schedule to ship the industry’s first 40-nm Stratix IV FPGAs [with up to 13.3M gates] later this year…hundreds of early adopter customers have already begun designs with these devices”
- John P. Daane, Altera’s CEO
Operating Highlights
- Altera is gaining market share and should continue to do so
- For the first time Altera believes it has architectural leadership and process technology leadership
- Book to bill was slightly below one at Q2 end, but has since accelerated over one
- Rockwell Collins, a top-tier government electronics contractor, recognized Altera as the number-one supplier among its more than 10,000 suppliers
- Cash = $1.2B / LT debt = $500M
- Quarterly Dividend = $.05
Guidance: Q3 2008
- Revenue: Flat to down 3%
- Gross Margin: 66.5 - 67.5%
Altera was very upbeat; this was their first conference call in some time where the economy was not one of the major discussion points. Altera is benefitting from a couple relatively new markets showing strong growth: Automotive customers are moving from prototype to production cycles and the government now accounts for 7-8% of revenue.
Xilinx is still winning the war, for now…
Xilinx was less upbeat, but also did not seem too concerned about the economy. Xilinx deflected questions comparing Altera’s growth with its own by discussing how Altera dominated the 130 nanometer process node, which is now peaking; even though Altera’s results appeared to be driven by the adoption of its 65nm and 90nm products. In the next couple years 65nm and 90nm products should begin peaking. Xilinx currently has the majority of market share at 90nm and it supplies 90% of the FPGA industry’s 65nm sales. Xilinx’s mid-term future looks very bright with its dominance at the 65 nanometer node. However, Altera’s momentum is making up for some of the initial market share loss due to the delay in their migration to these smaller process nodes.
Analysts are concerned that Altera beat Xilinx to 40nm. This is surprising based on Xilinx’s much faster migration to 65nm. A few years ago Altera delayed their product launches because of a re-design focusing on low power consumption by unveiling what they call programmable power. Since then Altera has definitely benefitted from this low power leadership. This time it appears Xilinx is doing a re-design, to focus on its platform strategy.
Xilinx’s Q1 2009 Results (released July 16, 2008)
Financial Highlights
Operating Highlights
- Gross margin = 63.8%, up from 63.4% last quarter and 62.2% in Q1 2008
- YTD has repurchased $350M shares
- Quarterly Dividend = $0.14 a share (up from $.12 last quarter and Q1 2008)
- Cash ($1.24B) + LT Investments ($607M) - Convertible ($1B) = $831M
Guidance: Q2 2009
- Revenues: up 1% to down 3% Q/Q
- Gross margin: 63 - 64%
For years Willem Roelandts, Xilinx’s former CEO now chairman, advocated FPGAs as a platform for 3rd party IP. Currently developers still get most of the IP (intellectual property) directly from the FPGA vendors or by building it themselves. Xilinx’s new CEO, Moshe Gavrielov, is beating the same drum of 3rd party IP. Soon after joining the company this year he articulated how it now costs $70 million to design a new ASSP, using the latest process node. This requires a billion dollar plus market for such an endeavor to be viable.
Xilinx believes they need to focus less on being a technology innovator in order to create market opportunities, but instead pursue a business platform leadership position. This strategy appears to make sense since the last couple years analysts have pondered why ASIC designs have not seen a mass migration to FPGAs as expected. Instead ASIC starts have migrated more to ASSP products. This creates a huge opportunity for platform FPGAs to take market share from mid to low volume ASSPs.
This is not a war of attrition
Altera’s strategy is clear. Leverage its low power leadership and relationship with TSM to create the most advanced FPGAs that can be migrated to its structured ASIC product. Xilinx strategy is yet to be fully unveiled, but it is focused on building a platform that vendors can use to create IP and end products that will compete directly with lower volume ASSPs and IP now designed for ASICs/ASSPs.
No longer is Xilinx strategy to focus primarily on technical innovation to create market opportunity, but instead to focus on customer requirements to fill a technological product void. This year Xilinx has re-structured the entire organization around creating vertical platforms (devices that basically compete directly with ASSPs and microcontrollers) and horizontal platforms (generic platform devices).
Such platforms for ASSP and microcontroller replacement will require using standard soft-cores within a powerful FPGA for ease of use by end product device developers and easy IP migration to newer chips. In many respects this strategy relies more on Xilinx’s ability to build a platform ecosystem of vendors and 3rd party IP providers than offering the most cutting edge performance. Unfortunately for investors, Xilinx has yet to announce exactly how it intends to accomplish the task of building this platform ecosystem.
Both companies can be successful. Neither company has interest in starting a price war. Each company has thousands of legacy customers that are not looking to switch vendors.
While I personally think this recession overall will be one of the most severe in decades I think it will be relatively minor for the technology industry; compared to the 2001 technology-led recession. The semiconductor industry is a leading indicator for the technology sector. In their conference calls both companies showed minimal concerns about the economy affecting their business. This is a reassuring sign for the technology sector.
Disclosure: I own a little more XLNX than ALTR. I will reconsider this allocation based on Xilinx’s platform and process node announcements later this year. I have investments in no other company mentioned.
