Arthur Goldberg: Big Blue and Beyond

IBM Veteran Enjoys New Career at Hal

"The enemy is not the other company offering a Unix solution but rather the monolithic alternatives being proposed."

ABOUT ARTHUR GOLDBERG

by Richard Cole

"There are no second acts in American lives," wrote F. Scott Fitzgerald. Some might agree, but don't try to convince Art Goldberg of that. Art is a notable example that there are indeed second acts--and more--in today's IT industry.

Art spent over two decades at IBM, then left and started a new career for himself in Silicon Valley. Currently he serves as executive vice president of Hal Computer Systems, Inc., a Fujitsu Ltd. company based in Campbell, CA and the home of the Halstation 300 series, the industry's first 64-bit Sparc workstation.

Art admits he's "never been shy" about expressing what he thinks about the computer industry and the companies in it. His remarks include straightforward opinions about his life at IBM, the open systems community and the computer industry in general. At the same time, his conversation reflects the kind of excitement and optimism that is perhaps essential for a man who has continually reinvented himself as an industry player.

"This is a fantastic business," he says and he asserts that the computing industry still attracts the brightest and most creative graduates coming out of the colleges and universities. "The challenge," he adds--as if commenting on his own career--"is to stay sufficiently flexible and to keep learning from the people who are entering the industry today."

Act One

The beginning of Art's career was straight out of Brooklyn: Williamsburg, in fact, a working-class neighborhood overlooking the East River in New York. Art grew up in a large, close-knit family, which included 36 cousins. He went to Stuyvesant High (the "best high school in the world," he claims) and then to Brooklyn College as a math major. After a brief stint as a graduate student at the University of Chicago, he decided not to pursue a doctorate in math. He returned to New York and began looking for a job.

He received several job offers and decided on a position as systems engineer with IBM. "It was a real simple decision," he recalls. "They offered me $50 a month more than the others. It turned out to be one of the luckiest decisions I ever made."

After a few months, Art was assigned to the Brookhaven National Laboratory at Brookhaven, Long Island. IBM had several computer systems installed at the lab. For the next three years, he worked with the researchers at Brookhaven on a variety of high-energy physics applications, including applications for what was at that time state-of-the-art, realtime data reduction processing on data from experimental devices. "The work was fascinating," he says. "I was among some of the leading research scientists in the world."

In 1971, Art became aware of an opening as the lead IBM account representative at Princeton University. At that time, Princeton had one of the most advanced computer environments in the world, and Art leaped at the opportunity. He was assigned to the Princeton account and worked on IBM 7000-class machines such as the 7044 and 7094. He also mentions setting up what was then a "supercomputer": an IBM 360 Model 91 which had a "massive" 4MB of core memory. He helped researchers with a number of projects, including weather modeling and high-energy plasma physics.

Management Stripes

Fascinating as it was, Art's experience with research and universities wasn't necessarily a plus at IBM. "There seemed to be the attitude that while universities and labs were an interesting, fascinating and tough environment in which to sell, it still wasn't the business world," he explains.

To broaden his experience, he left the campus and helped train new IBM sales reps in the Princeton area. The next year, he joined the finance industry marketing group of IBM, "one of the most fascinating jobs I ever had at the company," he says. He became heavily involved in product strategies and marketing plans for retail banking, and became an expert on electronic funds transfer systems.

At this point in his career, it was time to earn some "management stripes" in business, so he became a marketing manager in Buffalo, NY. Of course, he had to face the Buffalo winters. "The first year was awful," he says. "We saw 200 inches of snow." He adds that the other winters were not so bad: "only 60 inches or so."

After three years in Buffalo working in sales for the finance and manufacturing territories, he was moved to the White Plains, NY, area, just north of the city. He became director of marketing in a program with General Electric and helped grow IBM's revenues in this area to $165 million over a three-year period.

RISC Development

Art asserts that his proudest achievement in business was helping to move IBM into the open systems arena. Art worked on IBM's first RT workstations: 32-bit computers supported by RISC microprocessors and IBM's Unix variant, AIX. The RT machines were succeeded by the RS/6000 midrange system. Designed primarily for scientific and engineering applications at first, the RS/6000 has become a popular server for a range of business and Internet applications.

Developing the new machines involved tough technical problems, but Art says the highest hurdles were bureaucratic. Commenting on Louis Gerstner's recent address at UniForum '96 in which the IBM CEO made a pledge of allegiance to open systems, Art says, "Back in '85 and '86 when we were trying to move the company to AIX and Unix, I would have killed to have had an IBM executive stand up and make that kind of speech."

According to Art, open systems developers at IBM faced outright hostility from other divisions in the company. "Our greatest competition in bringing the Unix products to market wasn't Sun [Microsystems] or Convergent or the player du jour on the outside," he says. "It was competition inside the company from product groups invested in the status quo. These groups didn't want RISC products in particular or Unix products in general ever to see the light of day."

Frustrations with IBM

Art's concerns with IBM soon extended beyond the company's resistance to open systems. By the late 1980s, it was apparent to everyone in the industry that Big Blue was in trouble. Its mainframe business--historically the company's main source of revenue--was in decline, midrange sales were faltering and the personal computer market had been taken from under its nose by a number of PC-clone manufacturers.

It was equally obvious that IBM's top management wasn't up to the task of turning the company around. "I enjoyed going to work every day," says Art, "but I had serious concerns about how IBM was managed and the way people were being moved up the ranks."

These frustrations led Art to leave IBM in 1988. After 24 years, this was not an easy decision. Looking back, he comments that IBM afforded him opportunities and rewarded him in ways that he could not have envisioned "as a kid growing up in Brooklyn." He adds, "It's very hard to step away from that kind of secure environment, knowing that you are going to be well-compensated, well-provided for and have all your benefits taken care of unless you do something incredibly stupid."

Ironically, IBM helped cement his decision to leave by offering an attractive severance package--another example, says Art, of IBM's mismanagement at the time. "The severance packages were very poorly thought out," he says. Initially, the packages were conceived to weed out the dead wood, but because the offer was made to everyone, it allowed the best and the brightest to leave the company in droves for new positions while the less productive workers hung on to their jobs.

"I was a beneficiary of the program, of course, so I'm not complaining," says Art. "The packages allowed IBM to get revenues in line. But overall, it had an extremely detrimental effect on the company."

A California Pioneer

Art moved to California to serve as CEO of Atherton, Inc., a company involved in developing computer-aided software engineering (CASE) tools. His first experience at the helm of a computer company did not work out as planned.

Atherton turned out to have unforeseen problems. Art--and the board who hired him--thought that the company had a shippable product and strong revenues. In reality, the CASE product was at best an advanced prototype, revenues were far lower than expected and the company suffered from morale problems and dissension. Art had to refocus on getting the product up to commercial grade and redouble efforts to increase financial support. He was able to turn the company around but decided to leave after several years, coming to Hal in 1991.

At Hal, he has been in charge of diverse areas such as human resources and architecture. He is now focused on sales, marketing and customer service. Recent months have been exciting times for Hal Computer Systems, primarily because of the company's newest product: the Halstation 300 series. In addition to being the first 64-bit Sparc workstation, the Halstation offers binary compatibility with all existing 32-bit Sparc/Solaris applications. "That was biggest challenge that we have faced at Hal," says Art. "Namely, developing a system capable of offering new functions and facilities while maintaining the backward compatibility with the systems built by Sun and others based on the older 32-bit versions of the Sparc architecture." Apparently the company has met this challenge. It backs up the claim for 32-bit compatibility with a money-back guarantee that customers' Solaris applications will run on the Halstation.

Art also looks at the long term for Hal Computer Systems. "Hal is unique; we're a small company with the backing of one of the largest companies in the industry." This position, he explains, gives Hal the agility and flexibility of a startup company with the financial and technical strength of a major player. "In five years," he says, "I see us as the principle open systems development center for the Fujitsu family of companies. The product range will be broader and will include state-of-the-art workstations and servers."

Art adds that as an "evangelist" for open systems, he takes particular satisfaction in helping to bring Hal's Sparc-based products to market. "I really enjoy seeing what they can do for customers and how they are changing the way people do computing."

Art has no plans to move again soon. Still, when asked about the future, he says that he can conceive of perhaps another chapter. "I'd like to think that I'm still young enough to do this one more time," he admits. "So whenever I finish my tenure here [at Hal], I'd like to find a young, innovative company again, become involved in it and help build it to a success."

Spoken like a true pioneer.