How John Knickerbocker spent a career cultivating new technologies
In four decades at IBM, John Knickerbocker knows that sometimes the seeds of a good idea can take time to grow.
In four decades at IBM, John Knickerbocker knows that sometimes the seeds of a good idea can take time to grow.
IBM Distinguished Engineer John Knickerbocker has worked on next-generation semiconductor packaging technologies at IBM for over 42 years. Working in this area for this long, he’s not only seen the field change dramatically, but he’s also had his fair share of projects that didn’t pan out. Sometimes the tools needed for the work simply didn’t exist. Sometimes industry partners didn’t agree on a path forward. And other times, it turned out that a project seemed like a failure, but in fact laid the groundwork for new opportunities that didn’t become clear until years later. Regardless of outcomes, Knickerbocker and his colleagues have relied on their curiosity and industriousness to invent new processes out of necessity, changing the field of microelectronics as they went.
Knickerbocker’s work at IBM Research has focused on (but isn’t limited to) electronic packaging, including advances in fine-pitch and multi-chip integration, and heterogeneous integration technologies — all of which help power IBM’s core technologies of systems products, hybrid cloud computing, generative AI, and quantum computing. He has also worked on precision smartphone components, healthcare applications, and diagnostic sensors.
Along the way, Knickerbocker has helped pioneer technologies for assembling and handling sensitive microelectronic components, developing new ways to shrink the size of chips, substrates, and batteries, and increase chip-to-chip and chip-to-package interconnections. He's been involved with making more connections per unit area from a chip, stacking chips, increasing connections at the edge of chips, and handling fragile thinned silicon wafers, chips, and packaging components that are all a fraction of the thickness of a single hair. In the process, he helped author hundreds of patents and directly contributed to shipping over a billion micro-electric components.
Most recently, Knickerbocker and his colleagues unveiled an interconnect technology for co-packaged optics. This breakthrough brings fiber-optic connections to the edge of chips in data centers, which can reduce the energy costs of current generative AI computing by an estimated 50% compared to conventional electrical wiring. Co-packaged optics also increases the communication bandwidth among GPUs by 1,000% and supports longer GPU-to-GPU communication distances within a data center, with the potential to massively speed up large model training.
But these high-tech achievements grew out of low-tech beginnings. Knickerbocker grew up on a small farm in upstate New York, the sixth child out of eight. “We were all taught at an early age that you will help the family,” he said. “You will work on the farm — whatever is needed.” His older siblings taught him how to carry out the duties that kept the farm running. “We didn’t have a lot of money, but we had family,” recalled Knickerbocker. It didn’t hurt that he loved the challenge of learning new skills like planting the garden, maintaining cars, and repairing lawnmower engines.
Beyond household chores, Knickerbocker and his siblings were encouraged to earn additional money on their own. Their parents emphasized that they would go to college, and they would need to pay their own way through scholarships and working to save money.
Their home’s fertile soil also provided an opportunity to make money. Knickerbocker estimated his family vegetable garden was about 6 to 8 acres, and the family canned and froze home-grown vegetables to eat beyond the growing season. Some of their excess produce went to churches, family, and friends. But the kids also sold veggies to earn money for college.
“My brother and I were entrepreneurial from an early age,” he said, “going down the street with a little wagon, two bushel baskets in it full of tomatoes, corn, and squash to sell.”
To this day, Knickerbocker maintains a love of farming: Each year he grows over 100 tomato plants from seed, keeping his family, friends, and neighbors well fed with salads and sauces from July to November every year. And while farming remains a passion, his blossoming love for technology meant that academics were always at the forefront for him.
Knickerbocker’s skills in math and science made him stand out in school. In his first year as an undergraduate at the State University of New York (SUNY) Alfred, he spent his January term doing lab work on ceramics chemistry. By the time he was a sophomore, he was teaching the course — a role usually reserved for graduate students.
Upon graduating he stuck around SUNY Alfred for another year to do his masters, then went on to do a Ph.D. in materials science and engineering at the University of Illinois Urbana-Champaign. When he finished his postdoc year in 1983, he had plenty of job offers.
But he really wanted to work for IBM. IBM’s technical work with semiconductors and advanced packaging technology seemed a good fit for his skills and interests, plus he saw IBM as a leader in advanced research and development that was actually making its way to products — a factor that was especially important in Knickerbocker’s eyes. Plus, it didn’t hurt that he’d spoken to researchers who gave him the impression that IBMers were friendly, motivated, and team oriented.
“I didn’t see anything from IBM, though, so I called up someone who had come to speak when I was in college, and I asked if there were any jobs,” recalled Knickerbocker. He ended up receiving two different offers from IBM: one at IBM’s then-headquarters in Endicott, New York, and one at IBM Microelectronics in East Fishkill, New York. After some negotiating, he ended up in East Fishkill, where he started working on multi-layer ceramic packaging. “We had developed new technologies that didn’t exist at the time,” said Knickerbocker. “So it was new tools, new materials, new processes — everything needed to be compatible with the product application requirements.”
Since these beginnings, Knickerbocker has spent more than four decades at IBM — his entire career so far — including the past 20 years at IBM Research. But this path wasn’t inevitable.
Early on he’d considered being a medical doctor, among other things. “I thought about being a contractor, but I’d go broke because I’d take longer than estimated,” he said. “I’m too much of a perfectionist that way.” He still does construction as a hobby, with an impressive repertoire including designing and building garages, stone fireplaces, and homeroom additions. He’s done framing, plumbing, electrical, and roofing, and even added a floor, raising a house up a level. “Seeing the progress on these projects is enjoyable,” he said. “It’s a nice diversion from the day-to-day.” He’d also considered being a forest ranger and, of course, a farmer. But as a young man he decided none of those careers would be a great fit for him in the long run, so he chose the life of a semiconductor electronic packaging engineer.
Knickerbocker is first and foremost an engineer, but throughout his time at IBM he’s also become an expert at working with clients to make sure that what he’s making will meet their needs — not just a product that works, but one that can be scaled up to manufacturing quantities, adhere to a schedule, hit a specific cost point, and compete in the marketplace. And with the recent co-packaged optics breakthrough, the team’s announcement was as much about sharing results as it was about getting these client conversations rolling. After all, their input will affect where our collaborative research goes.
I want to make things that make a difference for people.
“One of the most valuable things I can do is talk to potential clients and understand what their needs are,” said Knickerbocker. For him, the value lies in moving from a concept, to demonstrators, to making real products. “I don’t want this to be a research project that ends as a research project, and you walk away. I want to make things that make a difference for people.” Not everything works out that way, he acknowledged, “but if you don’t work toward that goal, you’ll certainly never get there.”
Internal colleagues, too, can help build the necessary connections, said Knickerbocker, so his relationships with other IBMers are crucial to doing good work. Tech sales staff, for example, relay vital feedback from representatives at client companies. “We really want to know what would make a difference for them, what would make their product better than what their competitors offer,” Knickerbocker said. “It can be an opportunity for a great product for them down the road, and some new developments for us along the way.”
Since announcing the co-packaged optics breakthrough in December, Knickerbocker and his colleagues began meeting with potential clients to find out what they need from this technology. “You learn a lot talking to others,” he said. “You think you’ve got it all in hand, but really you don’t, because everyone is bringing different needs to the table, plus they want to be competitive.”
You learn a lot talking to others. You think you’ve got it all in hand, but really you don’t, because everyone is bringing different needs to the table, plus they want to be competitive.
Not everything works out, though, said Knickerbocker. In the early 2000s, he and his colleagues spent several years working on variable-focus electronic contact lenses with a client.
“We just weren’t firing on all cylinders, and we missed,” he recalled. The electronics, variable focus lens, and integration for the contact lens all worked, but the battery wasn’t to the client’s targeted specifications — specific energy levels with a targeted battery chemistry. Knickerbocker’s team followed up with an alternative: a low-cost, rechargeable battery that far exceeded the energy specification. “We made it successfully, with the thought that we might reset the program schedules,” he said. “Unfortunately, the research demonstration took a year and a half, and the program could not be reset at that time.”
And yet, this project wasn’t a failure. It took advantage of the team’s 3D semiconductor electronic technology and fine-pitch interconnection advancements, which had already been shipped and used for wireless cellphone products prior to when the electronic lens project started. The team added multiple complementary strategic inventions along the way, not just for that application, but for potential future applications, too. “Making components smaller, with higher bandwidth communication and energy efficiency, is a big deal for edge computing and other wired or wireless applications,” said Knickerbocker.
“We said if we created differentiation, we could start to think about other industry application enhancements: portable near-field recharge, wireless communication links, power delivery, and very small batteries,” he said. Their work on miniaturizing electronics and electronic packaging continues to be used for higher bandwidth interconnections between chips, as well as for packaging chips with a much smaller interconnection pitch.
Their advancements also included pioneering work on silicon interposers and packaging that could be shrunk into many more connections — many thousands of connections as opposed to a few hundred — by going to a tighter pitch. As part of their work, they developed the necessary equipment, materials, and processes to safely handle hair-thin silicon wafers with new bonding and de-bonding technologies, and put vertical connections into them with through-silicon vias (TSVs). By 2008, they had built and qualified for production their first TSVs for a device called a power amplifier, which became a crucial low-cost component in major smartphone brands like Apple and Samsung — eventually shipping billions of units.
Sometimes Knickerbocker’s work at IBM involves sowing seeds that don’t grow quite as planned.
In our first interview for this profile, I’d asked Knickerbocker for an example of something that didn’t work out, and he had told me about the contact lens project. It had been a great project with a great team collaboration, and many challenging technology advancements overcome. But it hadn’t kept 100% to schedule, which led to a disappointing program stop. Years had passed since the project ended, and what stuck with him wasn’t the scrapped program, but all the things he and his IBM Research colleagues had accomplished along the way.
Much like his passions at home, the triumph is just as much about the process as it is about the end product. After more than four decades at IBM, Knickerbocker is still just as driven by his curiosity and his eagerness to collaborate as when he first joined. His face lights up, a smile crinkling the corners of his eyes whenever he describes all that he and his colleagues have accomplished.
“I’m proud of many things in my career, but I’m most proud of working with my fellow workers here at IBM and with our research and development collaborators,” he reflected. “Very few things are done as an individual anymore, it’s just too complex — it takes many people with diverse skills working together to take innovations from concept into prototypes, and then scale them into manufacturing and making products.”