The room at the WAVES conference, an industry gathering focused on advanced packaging and thermal management, was full of engineers who make their living moving heat out of chips. When Intel Chief Executive Lip-Bu Tan took the stage, he had an investment to announce that fit the room’s theme: the company has put money into a startup that makes artificial diamond wafers designed to pull heat away from processors. Tan called the material the next generation of chip cooling, according to 36Kr’s report on his remarks.
The logic of the bet is straightforward, even if the technology sounds exotic. As chips pack more transistors and more power into the same footprint, the heat they generate has become a constraint on performance. Diamond conducts heat better than any common material, which makes it theoretically ideal for spreading heat away from a processor core. The startup Intel backed has spent years learning to grow synthetic diamond at wafer scale, a manufacturing problem that has defeated larger companies for decades. If the approach works, it could change how the hottest chips are packaged.
Tan’s appearance at a thermal management conference was itself a signal of how the CEO’s priorities have shifted. Since taking the top job at Intel, Tan has argued that the company’s future depends on two things: regaining credibility as a foundry, the business of making chips for other companies, and staying at the leading edge of the technologies that will define the next generation of computing. Thermal management, once an afterthought in chip design, has become one of those technologies, because the performance of AI accelerators is increasingly limited by how fast they can shed heat.
The same day, Intel announced the appointment of a new executive to run its foundry business. The company named Lee Seok-hee as executive vice president in charge of the foundry unit, a role that has been one of the most difficult jobs in the semiconductor industry. Intel’s foundry arm has lost money for years while trying to win business from the same companies that compete with Intel in its own markets, and its progress has been measured in design wins that have come slower than executives promised.
Lee’s appointment adds a new layer to an already complex story. He joins an organization that has restructured its leadership multiple times since announcing its foundry ambitions, and his mandate is to convert Intel’s manufacturing capacity into a credible alternative to TSMC. The foundry business is capital-intensive, unforgiving and slow: customers take years to qualify a new supplier, and they rarely switch once they commit. Intel’s challenge is not just building good chips but convincing the industry’s most demanding buyers that its process technology is worth the risk.
The two moves together describe Tan’s strategy in miniature: invest in the technologies that will give Intel’s chips an edge, and rebuild the manufacturing operation that will give Intel’s customers a reason to return. The diamond cooling bet is long-horizon — commercial deployment is years away, if it happens at all — while the foundry appointment is about the next two years of execution. Both are responses to the same underlying problem: Intel lost its manufacturing lead, and Tan is trying to buy it back with a combination of patience and urgency.
The markets have been skeptical. Intel’s stock has lagged the semiconductor sector through the AI boom, and investors have shown limited appetite for funding the company’s manufacturing buildout at the scale required to compete with TSMC and Samsung. The foundry business needs billions of dollars of new investment to reach competitive process nodes, and Intel’s own product business, while improving, has not generated the profits that would make that investment painless. Tan has said Intel will focus its capital on the most important bets, and the diamond investment is small enough to fit that description.
The broader question is whether the company’s transformation can outrun its balance sheet. Intel has already taken dramatic steps, including spinning off parts of its business and restructuring its manufacturing operations, but the semiconductor industry rewards scale and punishes hesitation. The foundry race is a marathon that Intel entered late, and every year of delay gives TSMC more time to deepen its lead. The new foundry chief’s job is to close a gap measured in years while managing expectations measured in quarters.
For the chip industry as a whole, Intel’s struggles have been a cautionary tale about the cost of losing a manufacturing edge. The company that once defined the leading edge of semiconductor production now watches its designs get made by others, and its attempts to return to the foundry business have demonstrated how hard it is to win trust back. The diamond cooling investment points to a different truth: Intel still has the engineering talent to identify the technologies that will matter next, even if it has yet to prove it can turn them into products.
Tan’s approach is to keep moving in several directions at once — new materials, new manufacturing leadership, new product plans — and hope that the pieces align before the money runs out. The WAVES conference gave him a stage for the long-term bet; the executive appointment gave him a manager for the short-term grind. Whether either move works will be measured in years, and in the meantime, the company’s shareholders are left to watch a transformation that Intel itself acknowledges is far from complete.
