Developments in AI, quantum computing and robotics are moving so fast that corporations must stay close to startups to avoid falling behind.
The head of Intel, Lip-Bu Tan, says corporate venture investing is not a side bet for companies but critical for ensuring a large organisation does not fall behind on frontier technology.
“[Corporate venture] has become the eyes and ears for me,” he told the audience at the GCVI Summit in Monterey, pointing to developments in quantum computing, frontier AI models and so-called “physical AI”.
For a company like Intel, which sits at the heart of the computing stack, that visibility is not optional. “The moment you lose touch you become isolated,” he warned, recalling how incumbents have previously missed major shifts such as mobile and artificial intelligence.
Those remarks shed light on Tan’s decision to make sure Intel Capital, its investment arm, has remained part of Intel. The company had explored spinning out the unit, one of the oldest and most active corporate venture units globally, but that plan was shelved after Tan’s appointment as chief executive, signalling a renewed emphasis on keeping venture activity closely tied to the parent company.
His argument is straightforward: corporate venture is about strategic awareness and speed. Tan urged corporate investors to “double, triple down”, framing them as critical actors in shaping the broader innovation ecosystem. At a time when technological cycles are accelerating, passive observation is no longer enough.
That message also reflects a wider shift in how large technology companies view innovation. Rather than relying solely on internal R&D, many are seeking closer, more systematic engagement with startups and venture capital — whether through dedicated investment arms or direct balance-sheet investments.
Read more:
For Intel, staying close to the startup ecosystem has become important as the semiconductor industry enters a phase in which progress is no longer driven purely by incremental improvements in silicon. Tan pointed to a series of emerging bottlenecks that will define the next era: power management, cooling technologies and high-speed connectivity between increasingly complex compute clusters.
As chip performance pushes beyond five gigahertz, traditional air cooling is proving insufficient. Alternatives such as liquid and immersion cooling are moving from niche to necessity. At the same time, heat dissipation and materials innovation — from advanced substrates to artificial diamonds — are critical areas of research.
Connectivity presents another constraint. As data centres scale, the limitations of copper-based interconnects are more apparent, prompting a shift to photonics to enable faster data transfer between processors. Meanwhile, the long-standing dominance of CMOS (complementary metal–oxide–semiconductor) technology is being challenged by alternative materials such as gallium nitride and silicon carbide, which offer potential gains in efficiency and performance.
These are not incremental adjustments but systemic changes — precisely the kind that large organisations have historically struggled to anticipate. Tan’s emphasis on maintaining close ties with startups and venture investors is therefore as much about organisational design as it is about technology.
“When a company is successful over time, it builds layers,” he noted, arguing for a more startup-like culture with faster decision-making and closer visibility into engineering challenges.
GCVI Summit Rewind — summaries of all the GCVI sessions are available here.
Maija Palmer
Maija Palmer is editor of Global Venturing and puts together the weekly email newsletter (sign up here for free).
