Intel CEO Lip-Bu Tan says Elon Musk’s chip ambitions reflect a much bigger problem facing the tech world: artificial intelligence is growing faster than the semiconductor industry can support.
During a recent appearance on the No Priors podcast, Tan shared his thoughts on working with Musk on the Terafab project, an ambitious effort tied to the massive chip demand expected from electric vehicles, robotics, AI systems, and even future space-based data centers. According to previous comments from Musk, Intel is expected to manufacture chips for the project using its 14A process technology.
Tan spoke highly of Musk, calling him one of the most impressive entrepreneurs of the century. More importantly, he said both of them agree on a critical point: the world’s chipmaking infrastructure is not expanding quickly enough to keep pace with the explosive growth of AI.
The Intel chief explained that the industry needs more capacity, better productivity, and greater efficiency if it wants to meet the demands coming from AI, autonomous systems, robotics, and advanced computing. In his view, the current semiconductor supply chain still has major gaps that must be solved.
Tan also described Musk as unconventional and highly detail-oriented. He said Musk constantly questions why things are done in a traditional way and pushes teams to examine every step of a process. Tan said he finds that approach refreshing, adding that different viewpoints can help both sides discover a better path forward.
That mindset lines up with Musk’s well-known “first principles” approach, where problems are broken down to their most basic elements before solutions are rebuilt from the ground up. For a project as large as Terafab, that kind of thinking could be important, especially as Musk’s companies are expected to need enormous volumes of semiconductors for cars, humanoid robots, AI hardware, and data infrastructure.
Tan’s comments also suggest that Musk is taking chip supply very seriously. As AI and robotics become more central to Tesla and other Musk-led ventures, securing long-term semiconductor production could become just as important as battery supply or manufacturing scale.
Beyond Musk, Tan also discussed how artificial intelligence is reshaping the entire semiconductor industry. He said AI may ultimately have a bigger impact than the internet, especially because it can improve efficiency across many parts of business and engineering.
In chip design, for example, AI tools can help reduce development time, improve timing analysis, lower costs, and automate repetitive tasks. These improvements could make semiconductor companies faster and more competitive, but they also increase demand for advanced processors, GPUs, CPUs, memory, and manufacturing resources.
However, Tan warned that the AI boom is creating serious bottlenecks. One of the biggest is power. AI data centers require enormous electricity supplies, and some regions simply do not have enough available power to support rapid expansion.
Another issue that receives less attention is helium. Tan pointed out that helium plays an important role in semiconductor manufacturing, including processes such as vapor deposition and cooling during etching. If demand continues to rise sharply, helium availability could become another pressure point for chip production.
The most visible shortage, however, remains memory. AI systems rely heavily on advanced memory technologies, and demand has surged as companies race to build larger data centers and more powerful AI models. Tan said memory is currently the biggest shortage in the market, with companies scrambling to secure supply.
Even when manufacturers decide to expand production, new capacity does not arrive quickly. Building out fabrication facilities and increasing output can take years. That applies not only to memory, but also to CPUs, GPUs, and other key semiconductor components.
As a result, Tan expects strong demand and higher prices to continue affecting the market. He noted that when production costs rise, companies often have to pass those costs on to customers.
The broader message from Intel’s CEO is clear: AI is not just a software revolution. It is putting intense pressure on the physical infrastructure behind modern computing, from power grids and chip factories to rare materials and advanced memory supply.
For Intel, the opportunity is significant. If the company can execute on advanced manufacturing technologies such as Intel 14A and attract major customers with large-scale chip needs, it could strengthen its position in the global semiconductor race.
For Musk, the Terafab project could become a major step toward securing the chips needed for the next generation of AI-driven vehicles, robots, and computing systems.
And for the wider technology industry, Tan’s warning highlights a growing reality: the future of artificial intelligence will depend not only on smarter algorithms, but also on whether the world can build enough semiconductors to power them.
