The AI chip shortage just met a billionaire mining magnate with a plan to rebuild America's semiconductor stack from scratch.
The Summary
- I-Pulse Inc., co-founded by mining billionaire Robert Friedland, secured $250 million in US government funding to develop semiconductor and pulsed-power technology
- The investment is part of America's push to reduce dependence on foreign chip supply chains, particularly from Taiwan and South Korea
- Friedland brings minerals expertise to a manufacturing problem — chips need rare earth elements, and he knows where they are
The Signal
Robert Friedland made his fortune finding copper and cobalt in places other people weren't looking. Now he's applying that same contrarian instinct to semiconductors. I-Pulse's $250 million government backing isn't just another CHIPS Act handout. It's a bet that the chip shortage isn't just a manufacturing problem — it's a materials problem.
The US currently produces about 12% of the world's semiconductors, down from 37% in 1990. Taiwan Semiconductor Manufacturing Company makes roughly 90% of the world's most advanced chips. Every AI training run, every autonomous vehicle, every edge computing node — all roads lead back to Taipei. That's not a supply chain. That's a single point of failure.
"The chip shortage isn't just a manufacturing problem — it's a materials problem."
Here's what makes I-Pulse different from the dozen other semiconductor startups chasing government money:
- Friedland controls access to critical minerals through his mining empire
- Pulsed-power technology could enable new chip architectures that require different material inputs
- Vertical integration from mine to fab potentially sidesteps current bottlenecks
The pulsed-power angle matters more than it sounds. Traditional chip manufacturing relies on continuous power systems that limit what materials you can use and how you can structure transistors. Pulsed systems — delivering energy in controlled bursts — could enable new semiconductor designs that work with more abundant materials. Less dependence on ultra-pure silicon. More options for substrate materials.
This is the kind of infrastructure play that takes seven years to pay off and twenty years to become obvious in hindsight. The agent economy needs chips the way the industrial revolution needed steel. You can't build it if you can't make it. Right now, America is trying to build a distributed AI future on a centralized semiconductor present.
The Implication
Watch where Friedland's mining operations expand over the next 18 months. That's your map of which materials I-Pulse thinks will matter for next-generation chips. If you're building AI infrastructure or betting on the agent economy, semiconductor independence isn't a political talking point. It's the difference between scaling and waiting in line.
The real question isn't whether America can build its own chips. It's whether we can build the supply chain that builds the chips. That starts at the mine, not the fab.