On Tesla's July 22 earnings call, an analyst asked Elon Musk the obvious question: does it eventually make sense to combine Tesla and SpaceX? Musk didn't say yes. He also didn't say no. He said there's "more and more overlap" between the two companies, and pointed straight at Terafab, the chip project the two are building together, calling it "really going to be a gigantic project." He said he couldn't get into "combining companies and that kind of thing" on an earnings call — that has to go through "the appropriate process."
Tesla then had its worst week in years, and SpaceX — public only since its June IPO — slid alongside it. By the implied math, that makes any near-term deal ugly: SpaceX would have to issue the large majority of its own stock to buy Tesla at current prices.
The bankers can fight over the ticker symbols.
The part that matters to anyone running a technical org is that the organizational merger already happened — quietly, weeks before this earnings call, at the only level that actually determines whether a project like Terafab works.
Terafab is Musk's plan to build semiconductor capacity, backed by Tesla and SpaceX — which absorbed xAI earlier this year and now runs it as its SpaceXAI division — with Intel signing on in April as the foundry partner, running its 14A process. It spans two sites: a smaller pilot fab at Tesla's Giga Texas in Austin, and a far larger high-volume facility in Grimes County, about 90 miles northeast. Musk originally pegged the effort at $20 billion. The Grimes County plant alone tells a bigger story. A property-tax-abatement filing there put initial investment at $55 billion, rising to as much as $119 billion across all phases.
None of that money solves the actual bottleneck, which showed up back in March in a single job posting. Tesla started advertising for a Technical Program Manager to own Terafab "end-to-end program scoping — including factory design/construction from concept through execution, ramp-up, and production readiness."
The requirements: 10-plus years in program management, at least five of them in semiconductor or high-tech manufacturing, with a track record running projects north of $100 million in capex. Read that list against Tesla's and SpaceX's actual businesses. Neither company builds cars or rockets in a way that produces people who've stood up a wafer fab from bare concrete. That skill set doesn't exist inside either org chart. It never needed to, until now.
So they didn't wait for the posting to fill itself. In June, Tesla hired Gary Jiang, a 17-year Intel veteran with deep manufacturing leadership, as Director, Terafab, the first named executive tied to the project. That's the actual merger event. A single, specific person, pulled by name out of the one company on earth that already runs the process Tesla needed, because a job posting alone was never going to produce a qualified applicant pool wide enough to matter.
"There's more and more overlap... [Terafab is] really going to be a gigantic project."
— Elon Musk, Tesla Q2 2026 earnings call, July 22, 2026. "More and more overlap" via Forbes; "gigantic project" via Electrek.
This is the pattern every CTO or VP of Engineering should be watching, independent of what happens to the tickers. When a company launches a genuinely new capability — manufacturing bolted onto a software company, hardware bolted onto a services company, whatever your version of "chip fab" is — the instinct is to route it through the normal hiring funnel. Post the req, screen the applicants, fill the seat.
That funnel was built for roles your company already knows how to evaluate. It breaks the moment the skill doesn't exist inside your walls, because there's no internal signal for what "good" looks like and no pipeline of people who've done it. Tesla figured that out fast: one generic posting, then a direct, named hire out of the competitor that actually has the bench. The org chart merged before the cap tables did, because that's the part with a deadline.
If you're standing up a capability your company has never staffed before, the mistake is treating it like a normal req. It isn't one. You need someone who can identify the handful of people who've actually done the specific thing, reach them directly, and make the case. The ones worth hiring aren't reading job boards. Go find them by name.
VC5 Consulting runs named, direct search for the roles that don't have an internal comp band yet — the ones where the applicant pool from a job posting is functionally zero and the only path in is knowing exactly who's done it before. If you're building a capability your company has never hired for, let's talk.