John Chen Plans Pyramid's Future

Hong Kong native switched to open systems with Unisys

Name: John Chen
Age: 40
            Place of Birth: Hong Kong
          Current Position: CEO, Pyramid Technology
  Time in Current Position: four months
     Years in the Industry: 16
             Car He Drives: 1990 Lexus LS 400
Favorite Non-Work Activity: Golf (19 handicap)
Prediction for the Future of Open Systems: Whatever we all end
     up using in commercial computing, it will be open, whether it's
     Unix or (Microsoft) NT. I only expect a handful of players in
     this IT world from both the platform and the software perspective.
     So there's going to be a lot of consolidation. We've got too
     many chips in the industry, too many versions of Unix, too many
     different database companies, too many application companies,
     and too many platform companies. There will be a major shakeout
     across the board.
On UniForum: UniForum needs to focus on the marketing role, and
     you always do. Stay away from products and focus on getting the
     Unix industry moving in the right marketing sense. In the arena
     with other environments like NT or proprietary, there's a lot
     of marketing of Unix that needs to be done because over time,
     the buyer of technology won't even care if it's Unix or not.
     To assure that Unix has a proper place, UniForum is the key.

By Don Dugdale

When Pyramid Technology Corp., San Jose, CA, announced the details of its acquisition by Germany's Siemens Nixdorf Information Systems at UniForum '95 in March, it was an important occasion for John Chen for another reason - it marked Chen's appointment as Pyramid's CEO.

For Pyramid, the merger meant the solid backing of the $65 billion Munich-based electronics giant Siemens AG in its corner, seemingly assuring Pyramid's future viability. For Siemens, it meant the acquisition of high-end Unix technology and a very useful U.S. market presence. For Chen, it meant another step up the ladder of a career that for the past 10 years has been focused on open systems.

Aimed for the Ivy League

Chen grew up in Hong Kong. At 17, his family sent him to the United States to attend a private school in New England as preparation for admission to an Ivy League university. "I realized that if you wanted to get into a really good U.S. university, the only way to do it was to come here, brush up on your English, and get into the system," Chen says. He received his bachelor's from Brown University and his M.S. from the California Institute of Technology, both in electrical engineering.

In 1979, Chen began a 13-year stint with Burroughs Corp., which became Unisys in 1987. For the first six years he worked at plants from Pasadena to San Diego, in design engineering, manufacturing, semiconductors, and as project manager for application software-all related to mainframes. In 1985, while Chen had responsibility for the company's Pasadena plant, Burroughs management made a decision to move from proprietary systems to Unix. "Although proprietary was profitable, we needed to switch to open systems," Chen says. "We started building a Unix team. In those days, we had the feeling of being a stepchild because the mainframe was still thriving and growing. Even in 1987, when Unisys was created, Unisys was doing extremely well off the mainframe base on both the Sperry and the Burroughs sides. You can imagine the road blocks in trying to start a new business called 'open systems'."

But the open systems concept had already taken hold in Chen's mind. "I was fascinated by using commodity type products like microprocessors, and standard operating systems and interfaces, to do commercial computing," he says. "I thought that had to be the wave of the future. It's cheaper, easier, and faster. You could integrate other people's work and add value to the system along the way without becoming obsolete." Chen definitely saw open systems as a new and better design for IT systems, then dominated by mainframes-a prediction he sticks to. "I see it as replacing the IT engines," he says. "Many years from today, the way businesses do computing is going to be based on these standardized components and nobody's going to talk about whether it's Unix or the other thing. Although we now have this Unix vs. NT situation, I do see Unix as being the major engine of business computing."

At Burroughs, Chen's Unix project was something like a "skunk works," bent on developing a new product. Chen had been interested in one of IBM's failed projects called RT-an experimental predecessor to the PC in which IBM tried to put its OS/370 mainframe operating system on a desktop machine. "I started an equivalent project at Burroughs, trying to emulate a proprietary architecture with Intel or Motorola chips," Chen says. "Basically, we were trying to move a mainframe architecture onto a microprocessor. By 1987, that evolved to become the Unix effort. In 1985-86, microprocessors were used as a controller of some kind, not a CPU. Microprocessors were not the in thing to put the software on."

As the Unix systems emerged, Burroughs merged with Sperry and soon after Unisys bought Convergent Technologies, San Jose, CA. Chen helped merge the technologies of the three companies, managing a division focused on high-end technology. By 1990, when Chen was named vice president and general manager for Unix operations, Unisys' open systems revenue had reached $600 million.

The Move to Pyramid

Still, Chen was frustrated with what he perceived as Unisys' lack of commitment to Unix. "It's a big company and there are various agendas," he says. "I really couldn't get the company to focus on supporting the activity. I was looking for a company smaller and more focused on Unix and on business computing through Unix." He landed a job at Pyramid as executive vice president in charge of operations-which included engineering, manufacturing, quality, and MIS. He took over customer services and marketing as chief operating officer in 1992. Then, two years ago, he added sales to his responsibilities when he was named president of the company. Last March he succeeded Richard Lussier as CEO, with Lussier remaining as chairman and taking on a broader role with Siemens.

Pyramid provides high-end, Unix-based enterprise servers, with about seven percent of 1994's worldwide market for commercial open systems servers according to International Data Corp. Always impressed with Pyramid's symmetric multiprocessing technology, Chen has tried to move the products into a massively parallel processing (MPP) environment and clustered technology for better reliability. "We made a lot of leaps and bounds in products, making them more robust for commercial applications," Chen says. In addition, he made a big change in sales strategy after annual sales dropped from $227 million to $192 million between 1991 and 1992. during that period, Pyramid lost big customers like AT&T, Olivetti, and part of Hyundai. OEMs used to account for 75 to 80 percent of Pyramid's sales and now 80 to 85 percent is direct selling. That shift was necessary because of an increasing trend toward commodity pricing and smaller profit margins. "If you have an OEM in the middle, you then have a two- to three-tier distribution," Chen says. "The profit couldn't be shared properly at all tiers, meaning that not everybody could make money. Therefore, that particular model collapsed, so we had to shift."

With about 3,500 systems installed worldwide, Pyramid pushes its systems' high performance and capacity as well as scalability and higher availability with clustered technology. "Our average selling price is $450,000, so I don't sell PCs or desktops," Chen says. However, now that Pyramid is part of Siemens, the picture has changed a little. Together, Siemens and Pyramid top the worldwide midrange ($100,000 to $1 million) Unix server category, according to IDC. Chen, who is responsible for all Siemens computer business in the United States, plans to launch midrange and lower-end systems as time goes on, but still in the server category. "I'm not going desktop at this point," he says. "The majority of my focus over the next couple of years is on the very high-end, mission-critical environment."

Joining Siemens

Pyramid's acquisition by Siemens earlier this year-in a $205 million stock purchase in addition to Siemens' previous 17.6-percent ownership of the company-provided benefits to both sides. Number one for Siemens, Chen reports, is the acquisition of Pyramid's technology. "Siemens has always been an OEM of our technology and when they saw our new-generation MPP operating system, that really fit into their global plan." Also, Siemens Nixdorf Information Systems, although thoroughly dominant in Germany and Eastern Europe, had little U.S. presence. "That actually causes them problems in selling outside the United States because the U.S. is famous for being the trend setter," Chen says. "So they need that kind of global reach."

Pyramid, on the other hand, needed more viability, in the sense of convincing potential customers that the company will be around for years to come. With $218 million in annual sales for the year ending last fall, Pyramid had trouble selling to Fortune 500 accounts and combating giants like Hewlett-Packard, AT&T GIS, and IBM. "With all the commercial accounts we want to win, the technology won't do it all by itself," Chen says. "People always try to beat us up by saying we're too small to be viable for the future. That problem is solved by being part of Siemens."

With viability apparently no longer a question, Chen sees only growth ahead for Pyramid. "Right now we have about a $275 million company," he says. "I'd like to see us be at least a half a billion dollars in a couple of years, and grow above the industry average for the next five years. The industry average is 20 percent, so I'd like to make sure we grow better than 25 percent. Right now we just put our heads down and get it done. We've got good sets of products with a growing marketplace, and I don't know what excuses I could use for not being successful."