John Daane
Altering Altera's Course

A large company doesn’t handle like a sports car. You don’t streak down the speedway of success sensing the track through tightly linked steering and suspension components with refined controls ready to respond instantly to your slightest subtle input. It’s more like sailing a large racing yacht into the wind. Every maneuver needs to be planned in advance with the crew carefully choreographed. Together, you work against the competition and the elements, trying to reach the next mark first without anyone being knocked overboard by the boom. You reach your destination obliquely through a series of angled maneuvers, never traveling in a straight line. When a competitor gets ahead and steals your wind, you need to turn and tack away, setting a new course, looking for your own clear air where you can fill your sails again, accelerate back up to speed, and position yourself to recapture the lead.

In November 2000, when John Daane took the helm at Altera, it was time to tack. The company had become complacent with their lead in CPLDs, and as the market winds shifted and boom turned to bust, Xilinx took a commanding lead in the new FPGA-centric programmable logic market. Daane immediately took action, bringing Altera’s efforts into focus on a small number of high-value projects. “Altera has excellent people,” says John. “I saw a huge number of projects underway and not enough energy on the key efforts that would make us successful. We cancelled a lot of programs and put our energy behind the few that were critical.”

Those critical projects included correcting the problems with Altera’s struggling design tool suite and subsequently readying the then-new Stratix family for launch. A much-improved version of Altera’s Quartus design tools was needed to make customers successful with any new future architecture. While the original Quartus suite had been ambitious, it was feature-heavy and performance-light and had been suffering from a lack of maturity since its launch. A new and improved version--Quartus II--made a huge leap forward in performance, reliability, and robustness and, from a design tool perspective, put the company back on its feet again.

Stratix was a new, vastly improved architecture that would be the basis for a number of future product lines. Outsiders gave the project little hope of shipping on time, but Daane knew better. “First, we had complete confidence in TSMC-- our partnership with them was so strong that we were part of their de-bug process. Second, and most importantly, is that there is a lot of talent at this company, and people are willing to work hard to get things done. There was still a fierce competitiveness in the staff despite the hard times they’d been through.” The company executed well and launched the product on schedule.

With two pivotal successes under its belt, John Daane’s new Altera wasn’t ready to rest on its laurels. “We believe we always need to be innovating,” continues Daane. “This is not a market where you can afford to sit back and say ‘We’ve won.’ You have to always be thinking about the next thing you can do to get ahead.” Altera did just that following on the success of Stratix and Quartus II with a string of game-changing innovations, including its revolutionary Nios soft-core processor and its partner, the problem-simplifying SOPC Builder tool. Nios set the early standard for FPGA-based embedded systems by providing a capable and flexible soft-core processor that could be used in a wide variety of applications. The SOPC Builder tool got rave reviews from early users for its ability to simplify the process of assembling a complex system-on-chip using Altera’s FPGA fabric with processor and peripheral IP.

Leveraging his LSI Logic experiences, Daane also saw an emerging market in the structured ASIC space where easy-to-design mask programmable devices with vastly reduced NRE could provide a “middle” alternative between high-end cell-based ASICs and FPGAs. The combination of structured ASIC technology with FPGA-based development, prototyping, and even early production would form a powerful solution unmatched in the industry. From this idea came Altera’s “HardCopy” program, which allows a customer to create a faster, cheaper, mask-programmed version of his design for final production after working the bugs out using FPGAs.

Even with the softness in the telecom market that began in 2001, Daane knew that the heyday of FPGA was not over. He felt that programmable logic represented a key technology that could solve a broad spectrum of problems, and he wanted to create solutions that could penetrate other higher-volume markets such as consumer, automotive, and industrial more effectively. In the process of tacking to gain a lead, one of his maneuvers was to get much closer to his customers, including getting them involved in product definitions. “Our customers made it clear that cost was a huge issue and was going to become increasingly more important as their markets became more competitive.” With device cost being one of the major barriers to FPGA deployment in high-volume production, Altera decided to create a new family of devices specifically cost-optimized from the ground up. In Sept of 2002 they announced their super-low-cost Cyclone family, which won almost immediate market acceptance. “Cyclone is now in the hands of over 3000 customers around the world –a majority of whom are in the high volume consumer digital market. “As of last quarter, we’d already shipped 2 million Cyclone devices and are on track to ship way over a million just this quarter alone. Cyclone has been a phenomenal success.”

Daane is a family man who finds time for few activities beyond his responsibilities at Altera and at home. Daane’s hobby of wine making reflects a patience and understanding of long-term strategy that one might not expect from the industry’s youngest CEO. Born and raised in Marin County, John graduated from UC Berkeley. His undergraduate job as a summer intern at LSI Logic gave him a jump start in the semiconductor industry. He spent 15 years there and rose quickly in the company before being tapped for the top spot at Altera. John’s shirt-sleeves-and-down-to-business approach was a bit of a departure from the previously stodgy Altera culture and quickly won over the staff. The secret of John’s leadership seems to be in the application of focused execution on high-value, innovative projects. People follow John’s lead because they trust his judgment and believe he’s leading the company on a path to success.

Now, over three years into Daane’s tenure, Altera has become a case study in strategic execution. This seems to be the year of the sequel at Altera with the launch of new, highly competitive versions of previously successful product lines including Stratix II, Nios II, Max II, and an expected announcement of a successor to the low-cost Cyclone line as well as the HardCopy Stratix structured ASIC option. Altera’s Quartus II tool set has been undergoing continuous innovation as well, with FPGA Journal’s latest market survey showing it leading the pack in customer satisfaction among FPGA vendor-supplied tool offerings.

Daane is equally adept in both the marketing and technology sides of the house, and he has led a revolution in customer-driven rather than technology-driven products at Altera. Nowhere is this clearer than with the recent announcement of the Max II CPLD line, which turned the technology map sideways by using an FPGA architecture to deliver a product to CPLD designers. Rather than focusing on a logical, inertial extension of the CPLD architecture, Altera took a market-driven approach and studied their CPLD customer base to find out what features and capabilities were most important. They then delivered those features with the most capable architecture available, which turned out to be an LUT-based FPGA architecture rather than a CPLD structure. Following the theme of giving the customers what they want, however, Altera markets the product as a CPLD with CPLD pricing and a CPLD design flow.

True to his own mantra, Daane is not declaring victory with the recovery he’s engineered thus far at Altera. “As soon as you quit innovating, somebody comes up and changes the rules without you,” Daane says. “You have to always remain paranoid and always be coming up with the next idea.” Altera has a culture of competitiveness, and that vision resonates well with the team.

Daane is upbeat about the future of programmable logic. “Programmable logic products today have the ability not only to make a big dent in the ASIC space, but also to take on functions normally offered by standalone DSPCs, microcontrollers, microprocessors and memory chips. With the market moving so fast and more and more competitors jumping into the field, more and more system designers are looking to PLDs as their competitive edge. This is the era of programmable logic.”

In yacht racing, when a competitor feels threatened by a challenger, he often adopts a conservative strategy known as “covering,” where the leader mirrors the moves of the challenger, trying to maintain his lead by preventing the challenger from getting a tactical advantage or catching a favorable wind shift. Every time the challenger tacks, the leader turns to follow. The net result is that the challenger becomes the innovator, and the leader becomes the follower. With John Daane steering Altera, and their sights set on the number one spot, they have taken on this innovator role, and the pace and precision with which they make each move poses a serious challenge to any competitor.

Kevin Morris, FPGA and Programmable Logic Journal

May 25, 2004

[back to top]

Comments on this article? Send them to comments@fpgajournal.com