Newtown Researcher Honored For Semiconductor Work

By John Voket

As a scientist working with GE in the 1950s and early 60s, Ken Robdell’s father was responsible for helping develop the magnetic analog tape. That invention, the precursor to 8-tracks and cassettes, quickly revolutionized the way people listened to and stored music, as well as many other types of information.

Over the past 20 years, the junior Mr Robdell has been working to revolutionize a new generation of technology — computer processors and semiconductors. And anyone who works on a laptop or plays high-end video games owes a debt of gratitude to this brilliant Newtown resident and his research teammates at IBM’s T.J. Watson Research Center.

Mr Robdell and several members of his team at IBM were recently named among 2006 Inventors of the Year by the New York Intellectual Property Law Association for a research patent they authored on a complex metallurgy used to make computer processors and semiconductors faster and more efficient. According to a release from the association, this patent was recognized as a seminal invention that enabled copper interconnects (wiring) to be used in semiconductors.

Because this invention was so easy to replicate, the process has been adopted by all global semiconductor companies that have switched from aluminum to copper interconnects. In an attempt to put an extremely complex engineering and scientific activity into perspective, it may be compared to changing every roadway in Connecticut from dirt to pavement, with intersections that allow traffic to flow without interruption at the highest possible speed — then putting it all onto a piece of compressed sand about as big as a postage stamp.

According to Mr Robdell, the road leading to his latest development, and his recent honor, was a challenging one.

“This current project took almost ten years to bring to fruition,” Mr Robdell said. “But once our team was convinced it could work, it took almost three years to convince our own manufacturing people.”

He described having to convince a room full of manufacturing engineers and IBM executives to adopt “a process completely adverse to everything they knew up to that moment.”

“We were asking them to actually dip silicon wafers into liquid metal, then to grind down the surface with a polisher that was used to smooth the surfaces of windows on 747 jet airliners,” Mr Robdell said. “And we were asking them to do it with copper.”

Mr Robdell explained that up to that time, copper was basically regarded as a contaminant in computer chip manufacturing. But he and his team were convinced that using copper integration and a finishing process based on a centuries-old jewelry making technique could change the world of computers forever.

“Once we were able to convince enough of them that our idea would work, it eventually led to IBM’s first microprocessor, which was referred to as the power PC chip,” he said. “It offered a tremendous speed advantage over anything we had developed up to that time.”

The idea for that power PC chip was subsequently sold by IBM and a few generations later began showing up in both early generation Macintosh computers and in PCs as the Intel processor. But that did not stop Mr Robdell’s team from continuing to refine and improve the process for making these semiconductors even faster and better.

The patents for many of the processes used in manufacturing were still retained by IBM, and licensed to manufacturing vendors worldwide, creating a huge return for the company in subsequent years.

More recently, Mr Robdell and his team not only worked on refining the microprocessor manufacturing process to make the chips faster, but to put a lot more processing mechanisms on increasingly smaller chips. This is what led to his eventual Inventor of the Year honor.

“This patent has to do with filling the really fine features,” he said. “We already knew we could keep the copper contaminants out of the silicon, but now we are filling the chip with structures so small, you could get thousands of them into a space the diameter of a human hair.”

Although the assembly and engineering processes Mr Robdell’s team developed have been working effectively since 1997, their continued refining has brought a phenomenal improvement in terms of the speed in which computer information is transferred within the processor — the powerplant of the computer’s brain.

“Today we see the latest generation of that elementary power PC chip in almost every computer game system like Nintendo and X-Box, in laptops and PC systems and servers,” he said. “But I didn’t invent it. I got the Inventor of the Year honor for the patent, which is shared by the ten original authors.”

He smiled thinking back ten years when the industry was getting excited about very expensive computers featuring a 256 mghz processor.

“Today, a two to three gig processor is common in laptops that cost under $1,000 and video game systems that go for a couple hundred dollars,” he said. “But these processors also allow more work to get done faster, using lower power. It’s not uncommon to run several types of software on a notebook-sized laptop for four to six hours on a small lithium battery.”

Looking back even before his work with IBM, Mr Robdell wondered where his life might be today if he had not followed in his father’s footsteps. In fact, if he had taken his mother’s advice and pursued a college program in liberal arts, this invention and the world of computing and video gaming might not exist today as we know it.

And although he works with very complex engineering, having authored or co-authored almost 60 patents in his 20 years in the field, Mr Robdell still enjoys talking about science in a way that helps even the most uninterested subjects understand and appreciate the benefits of his calling.

“I really like coming into the schools and talking with kids about science in a way that helps them understand how important this field is. And maybe a few of them will decide to pursue it,” he said. “I often find out in the elementary and high school settings, the understanding and appreciation of the physical sciences is really bad.

“And that’s too bad because the physical sciences are really fascinating,” he added.