IBM Research – Zurich, a Success Story

Chris Sciacca and Christophe Rossel

 

Fig. 1. The building of IBM Research in Adliswil in 1956. Inset: Ambros Speiser, first lab director, in discussion with Thomas Watson Jr., CEO of IBM
Fig. 2. Hans Peter Louis one of the earlier research staff members in the 50’s, in front of a prototype technology called the phototron, which was never finished.

An Industrial Lab Grows in Switzerland

As the Zurich Lab recognizes its 50th year in the leafy Zurich suburb of Rüschlikon, a closer look inspects the details of how this lab became the home of four Nobel Prize laureates and countless innovations spanning material sciences, communications, analytics and Big Data.

There are several reasons why IBM was considering a research lab outside of the United States in the 1950s. At this time IBM was in its heyday. The company was financially strong, and the success of the recently opened San Jose lab made the management realize the benefits of having research conducted with the support of headquarters in New York, but without the local stress and peer pressure.

Switzerland wasn’t IBM’s first option for a European research lab; London and Amsterdam were also on the short list, and in 1955 an IBM electrical engineer named Arthur Samuel was tasked with scouting the three cities.

While officials in England were receptive to the idea, the proposed location was in the suburbs of London that he described as, "the most dismal places that I have ever seen." And shortly after, Samuel passed on England and travelled to Switzerland, to a completely different experience - he never made it on to Amsterdam.

Simultaneously, as Samuel was visiting Switzerland, Ambros Speiser, a young Swiss electrical engineer from ETH Zurich, had applied for positions at Remington Rand and IBM. He never got a response from Rand, but by that summer Speiser became an IBMer.

How to Build a Research Lab

Now that Speiser was on board he was tasked with building a new laboratory, and the challenges he faced were immense.

As he tells it in the IEEE Annals of the History of Computing [1], "There was no established pattern to follow - an industrial laboratory, separate from production facilities, did not exist in Switzerland."

But Speiser knew he needed to be close to Zurich, its universities and within reach of public transport. After viewing a number of locations, he decided to rent the wing of a Swiss stationary company in Adliswil, which was at the end of a tram line and only a few kilometers from the city.

Having the building for the new lab, Speiser needed brilliant scientists and engineers. Leveraging his acquaintances, professional societies and contacts at ETH, he began a recruiting campaign and quickly amassed a team from all parts of Europe.
The goal set by IBM management was to build new and better computer hardware. At the time everyone knew that vacuum tubes would be replaced by solid-state circuits, so it was obvious that IBM should begin developing transistors and magnetic devices, but this was never formalized.

This lack of direction weighed heavily on Speiser’s mind because he knew that similar research was being conducted inside and outside of IBM and that they would never achieve the global recognition he so desired by doing the same science as everyone else.

After building a stronger rapport with Research’s new management team in 1958, Speiser was given new direction for the Zurich Lab to change from electronics to physics, with a focus on solid-state as the basis for electronic devices of the future.

Once again Speiser went on recruiting missions across Europe to find young, creative physicists. Little did he know at the time that he was also laying the groundwork for what would become a renowned team for decades to come.

Fig. 3. Introduced in 1956, the IBM 305 RAMAC (Random Access Memory Accounting System) was an electronic general purpose data-processing machine that maintained business records on a real-time basis. The 305 RAMAC was one of the last vacuum tube systems designed by IBM, and more than 1000 of them were built before production ended in 1961.

Moving to Rüschlikon

In addition to electrical engineering and physics, Speiser soon added a mathematics department, and quickly the lab was outgrowing its modest space in Adliswil. Knowing that the growth would continue and to establish a more stable reputation, he requested approval from Thomas Watson Jr., IBM’s CEO at the time and son of the founder, to look for a new location where the lab would have its own facilities.

Again, Speiser knew the new lab had to be close to the city, the airport, but most importantly to ETH Zurich, and after some debate a 10-acre site was purchased in Rüschlikon for $400,000.

As for the actual design of the lab, Speiser’s main tenant was that it created spaces for personal interaction, which he referred to as "a vital process for a research laboratory". To assure this idea, instead of building up, he wanted the lab to be horizontal with long corridors to encourage chance meetings, similar to the design of Bell Laboratories in New Jersey. It was also important for the lab to have a proper cafeteria for informal discussion and an auditorium for guest lectures. While initially this was met with some skepticism because of its costs, he eventually got his wish and construction started in 1961. The lab was officially inaugurated in front of several hundred guests on 23 May 1963, that is, 50 years ago.

Impacting the Future of the Lab

In the coming decade, the groundwork of Speiser and his successor began to bear fruit. The idea to start a physics department resulted in the hiring of Heinrich Rohrer and Karl Alex Müller, who through their research and subsequent publications began to cement a strong reputation for the Zurich Lab, which reached its apex in the mid-80s when these two scientists and their colleagues Gerd Binnig and Georg Bednorz were recognized with the Nobel Prize for Physics in 1986 and 1987.

This period also saw the addition of a new line of research in communications. Once again the lab succeeded with the development of the token ring and trellis-coded modulation, each playing a critical role in making the Internet what it is today.

By 1987 the lab also had its own manufacturing line for semiconductor lasers which were used by telecommunication equipment manufacturers. The research became so successful that the Uniphase Corporation acquired the technology and the people from IBM for $45 million, a huge sum for the future telecommunications giant with only 500 people at the time.

The Crash and Recovery

With the end of the 80s, towards the mid-90s, IBM found itself in a dire position. IBM’s near-death experience was caused by its failure to recognize that the 40-year-old mainframe computing model was out of touch with the needs of clients, and other firms like Sun Microsystems pounced on the opportunity.

This experience also impacted IBM Research. Throughout the 1970s IBM Research was corporate funded, it had its own research agenda and occasionally it did some technology transfer, but it was not done in a very coordinated manner because the funding kept coming every year and the profit margins were strong. This afforded the scientists the freedom they needed.

By the 1980s IBM began doing more applied research, and management took a more active role in influencing the direction in which the developments had to go. For example, IBM started joint programs between Research and the product divisions with a shared agenda that both parties, Research and Development, had to agree upon. IBM also created collaborative teams to accelerate the transfer of research results which went into products spanning from storage to personal computers. It’s strange to look back at this now, as today this seems so obvious.

In the 1990s change truly came. To preserve Research, scientists in Zurich tried to become more proactive in working on actual customer problems. At the time this was unheard of at IBM and a large reason why the company stumbled. The idea was to interact with clients, gain insights into their challenges, and find solutions. The concept was a great success, and in 2000 the Zurich Lab opened up a dedicated facility called the Industry Solutions Lab (ISL), with the goal of hosting and interacting with clients on a daily basis.

Today, there are similar facilities around the world hosting hundreds of clients every month and working directly with clients. This is a fundamental strategy across all twelve IBM Research labs on the six continents. "The world is now our lab," as says Dr. John E. Kelly III, IBM senior vice president and director of Research.

Fig. 4. The newly built Binnig and Rohrer Nanotechnology Center and as inset an example of the complete atomic structure of a pentacene molecule resolved by AFM [2]. The extreme resolution of the C, H atoms and chemical bonds is achieved by the CO molecule attached to the tip.

IBM Research in Zurich Today

Under the direction of seven successive lab directors, the expansion in Zurich continued well into 2000s. Today, there are five departments, namely, storage, computer science and systems in addition to physics (science and technology) and mathematics (mathematics and computational sciences).

In addition, the lab has a new cutting-edge facility called the Binnig and Rohrer Nanotechnology Center, named for the two Nobel Laureates. When the current lab director requested the funding to upgrade the existing clean rooms on the campus he was greeted with a pleasant surprise: "I was told to make it much bigger and to find a partner. ETH Zurich was an obvious and logical choice," says Dr. Matthias Kaiserswerth, the current director of IBM Research - Zurich.

No one could have predicted it, but Speiser’s intuition to keep the lab close to ETH Zurich was a fortuitous decision. Nobel Laureates K. Alex Müller, Georg Bednorz and Heinrich Rohrer all came from ETH. And now nearly 60 years later, the partners built a $90 million facility, which features a large clean room and in particular six Noise Free Labs unlike any in the world.

Outside of the nano world, IBM scientists are working on some of the greatest challenges of our society today.

On Earth Day 2013, scientists in Zurich announced that they will be building an affordable photovoltaic system capable of concentrating solar radiation 2,000 times and converting 80 percent of the incoming radiation into useful energy. The system can also provide desalinated water and cool air in sunny, remote locations where both are often in short supply.

Another team is collaborating with a consortium of scientists in the Netherlands and South Africa on extremely fast, but low-power exascale computer systems aimed at developing advanced technologies for handling the Big Data that will be produced by the Square Kilometer Array (SKA), the world’s largest and most sensitive radio telescope that consortium will build.

And to improve the much-strained energy grid, IBM scientists are collaborating with utility companies in Denmark, Austria and Switzerland to improve to balance between demand and the supply of renewable energy.

While much has changed at IBM Research – Zurich, the essence of collaboration and the spirit of innovation and excellence that Speiser envisioned remains true to this day.

 

[1] A. P. Speiser, IEEE Annals of the History of Computing 20(1), 15 (Jan.-March 1998).
[2] L. Gross, F. Mohn, N. Moll, P. Liljeroth, G. Meyer, Science 325, 1110 (2009).

 

 

[Released: July 2013]