One thing that US policymakers seem to agree on, despite many other differences, is that the country must lead technologically to maintain a position of economic and geopolitical dominance. How to provide such leadership will be a critical question for the next US president and his staff.
The last two administrations have taken extraordinary steps to maintain an advantage in both chipmaking and AI, two fields that are inextricably and intricately linked. The U.S. and its allies have restricted the export of cutting-edge chips and silicon-making equipment to key geopolitical rivals (namely China). In 2022, the U.S. also passed the CHIPS Act, which would allocate $280 billion to bring more chipmaking back to American soil.
Laurie E. Locascioundersecretary for standards and technology at the Commerce Department and director of the National Institute of Standards and Technology, helps oversee government investment in chips. He tells WIRED that inventing up-to-date chip designs and manufacturing techniques is key to ensuring the U.S.’s technological dominance in AI. He adds that chip packaging—the process of combining components in up-to-date ways to enhance performance—could be particularly significant for the next wave of AI.
Locascio recently sat down with WIRED senior writer Will Knight at the Commerce Department in Washington. Their conversation has been lightly edited for length and clarity.
How have generative AI and ChatGPT changed the US government’s microprocessor priorities?
During Covid, we couldn’t get the basic chips, the technologies that we rely on for everything. But now the conversation is changing. People are realizing that we need the most advanced chips. We’re really at the top of our game in AI, and AI is changing the game for a lot of companies. So the most significant thing on everyone’s mind right now is AI chips.
What does this actually mean in the context of the CHIPS Act?
We don’t just want to bring today’s technology to our shores. We really need to continue that, being able to produce the next generation and the next innovations that come out of the labs. We’ve always been really at the forefront of innovation and creativity in this space, so that’s our advantage.
That’s why the CHIPS Act has these two components — $11 billion for research and development and $39 billion for manufacturing. Those two have to work in sync with each other because it’s our ability to innovate that’s going to make these manufacturers want to stay here. So we’re developing, in sync with the research and development community, up-to-date types of technologies that can be put directly on the production line.
Faster AI chips are key to AI companies’ efforts to build more powerful AI. How is this need shaping investments in next-generation manufacturing?
I would say it has really focused some of our thinking in certain areas. For example, when we think about how we spend $3 billion in “advanced packaging” we’re really thinking about the AI problem now. We understand how significant packaging can be to this particular set of problems. I would say it’s really focused some of our thinking in certain areas both on the manufacturing side and on the R&D side.
Why is packaging, i.e. combining different elements, so significant?
It may sound like a cliché, but packaging is enabling the development of 3D chip architectures that will truly accelerate the power of AI chips and support build the AI revolution. We just announced Notice of Intent to Fund $1.6 Billion. It focuses on many areas, but the power and thermal requirements associated with AI systems are really significant.