UpTrajectory Review

Elon Musk has floated the idea that his companies could eventually design and manufacture their own semiconductors, telling an audience that Tesla and SpaceX would probably figure out how to make chips better than anyone thanks to their experience building specialized equipment in-house. The remark, thin on detail but heavy on ambition, fits a pattern: Musk has repeatedly pulled critical components into his own orbit rather than rely on external suppliers. Tesla already designs the AI training chips that power its autonomous-driving program, and SpaceX builds roughly three-quarters of its rocket hardware internally. The claim here is not that Musk will build a fab next year; it is that vertical integration, once a strategy reserved for industrial giants with patient capital, is now being framed as a competitive necessity even in capital-hungry categories like silicon.

For a small-business operator, the instinct is to dismiss this as a billionaire's vanity project. That would be a mistake. Musk's bet is a signal about where supplier power is heading. When the largest buyers in a market start designing their own inputs, it usually means they expect the middle layer, the merchant suppliers, to get squeezed on price, capacity, or priority. If you run a business that depends on custom hardware, specialized components, or any outsourced process that a larger player might internalize, the lesson is not to copy Musk but to audit your own dependencies. Which of your vendors could decide you are no longer a priority customer? Which inputs could a well-capitalized competitor bring in-house and use to undercut you on cost or speed?

What is genuinely contested here is whether Musk's confidence is justified. Designing chips is one thing; manufacturing them at scale is a discipline that has humbled Intel, the company that literally invented the modern semiconductor industry. Tesla's Dojo program has already faced delays and leadership turnover, and the foundry economics that make TSMC dominant are not easily replicated by a company whose core competency is assembling vehicles and launching rockets. We are skeptical that Musk's companies will out-manufacture TSMC or Samsung. But we think the strategic logic, controlling the silicon that runs your software, is sound and increasingly mainstream. Amazon, Google, and Apple have all reached the same conclusion. Musk is late to the design table, even if he talks like an early arrival.

The second-order effects matter more than the headline. If large buyers continue to absorb chip design talent and foundry capacity for internal projects, small companies that need custom silicon or even priority access to standard components could face longer lead times and higher NRE costs. Design houses that once served a broad market may retool to serve a few deep-pocketed anchor clients. On the flip side, the tooling for chip design has never been more accessible; open-source instruction sets and cloud-based design tools mean a small team can prototype an ASIC without a billion-dollar budget. The gap is not in designing a chip but in getting it fabricated affordably at low volume, and that is where merchant foundries and aggregators still earn their margin.

What to watch: whether Tesla actually expands beyond AI-training silicon into general-purpose or automotive-grade manufacturing, and whether SpaceX's Starlink unit begins designing its own modem chips at volume. Both would signal that the in-house trend is accelerating past the hyperscalers and into adjacent hardware categories. In the meantime, the practical move for a small operator is to map your supply chain's single points of failure and ask whether any of your critical vendors are one vertical-integration announcement away from deprioritizing you. If the answer is yes, dual-source now, while the merchant market still wants your business.

“figure out how to make chips better than anyone” — Bloomberg Businessweek

Takeaway: Audit which of your critical suppliers a larger competitor could bring in-house; dual-source those inputs before merchant capacity tightens.