Semiconductors power a wide range of electronics, from cellphones and computers to cars, fighter jets, and medical devices. A global chip shortage that began in 2020 upended the automotive industry, drove up prices of electronics such as televisions, and exposed supply chain vulnerabilities.
Now it’s time to focus on strengthening the semiconductor supply chain and moving chip manufacturing back to the U.S., said Pat Gelsinger, CEO of Intel, the largest chip manufacturer in the U.S.
“Everything digital runs on semiconductors,” Gelsinger said during a March visit to MIT as part of the Manufacturing@MIT Distinguished Speaker Series. “Where do you want that digital future to be built?”
Gelsinger joined Intel in 1979, eventually becoming the company’s chief technology officer. He left in 2009 to become president and chief operating officer of EMC, which later merged with Dell, before rejoining Intel in 2021 as CEO.
During his talk, Gelsinger discussed why he supports the CHIPS Act, which aims to reestablish chip manufacturing in the U.S.; where he sees weaknesses in existing semiconductor supply chains; and why he’s optimistic about advancements in chip technology.
The value of reshoring and the CHIPS Act
The CHIPS Act, which was signed into law in August 2022, provides $52 billion in funding for American companies to create and expand chip manufacturing facilities. Gelsinger called the act the “most significant piece of industrial policy legislation in the United States since World War II.”
“Geopolitics has been defined by oil over the last 50 years,” he said. “Technology supply chains are more important for a digital future than oil for the next 50 years.”
The act is an essential part of reducing American reliance on other countries for chips, Gelsinger said. In 1990, 80% of semiconductor chips were built in the U.S. and Europe, he noted, but today 80% of semiconductors are built in Asia, with a large concentration in Taiwan.
This is concerning because Taiwan is geopolitically unstable, lacks some natural resources and energy reserves, and is located on a major fault line, Gelsinger said. “You have this extraordinary world dependency on a very small area of the planet,” he said. “This is not good for the geopolitical stability [or] the resilience of our supply chains.”
Gelsinger said he is confident that the U.S. will “invent” the future. “The question in my mind is, will we manufacture the future?” he said. “Are we going to own the supply chain of manufacturing and deliver this at scale?”
Intel has committed to expanding semiconductor manufacturing in the U.S., including spending an initial $20 billion on new chip fabrication plants in Ohio and another $20 billion on fabrication plants in Arizona. Gelsinger said chip manufacturing could turn the Rust Belt into the “Silicon Heartland.”
“What we’re doing with these projects is reshoring, rebalancing our manufacturing, leading with the core technology for the digital future,” he said. “What I believe is essentially the future of the planet rests on what we are doing in building our technology a future.”
Building the workforce of the future
Those fabrication plants need a workforce.
“I need a class of workers that are going to be able to create the most efficient factories with the highest automation, metrology, tool efficiency, etc.,” Gelsinger said.
Intel has announced a $100 million education and workforce development plan, with $50 million in funding from Intel and $50 million from the National Science Foundation. The workforce development initiative will focus on the Ohio plants.
With higher costs of labor in the United States, Gelsinger said he also needs to focus on being cost-effective rather than simply replicating the same jobs from overseas in the U.S. “We have to have better jobs using higher automation, more robotics, more innovation, AI efficiency,” he said.
Chips will continue to evolve
Moore’s law — the idea that the number of components on an integrated circuit doubles every two years — is in no danger of going away, Gelsinger said. The theory was named after Intel co-founder Gordon Moore, who died in late March, shortly after Gelsinger’s appearance at MIT.
But continuing to innovate at that pace is not easy. “We’re no longer in the golden era of Moore’s law,” Gelsinger said. “It’s much, much harder now. We’re probably doubling, effectively, closer to every three years now. So we’ve definitely seen a slowing; we’ve seen a shift in the economics associated with Moore’s law.”
The most advanced chips built today have about 100 billion transistors. Gelsinger said he thinks Intel can create a chip with a trillion transistors by the end of the decade — effectively doubling transistor density every two years.
Gelsinger said his optimism is based on several factors, including its new RibbonFET transistor and power delivery mechanism. Intel has also invested in cutting-edge lithography technology, he said, including high numerical aperture lithography, which will be used at Intel’s Oregon facility by the end of the year.
Finally, chips themselves are going from 2D to 3D, he said. “We’re stacking and being able to essentially innovate in three dimensions as opposed to two dimensions,” he said.
“Semiconductors is a volume manufacturing game,” Gelsinger said. “The engineering, the innovation is never done. If we are going to be an innovator at the heart and soul of the digital future, we must be a manufacturer at scale.”