Tesla Recruits Intel 18A Veterans to Lead Terafab Chip Facility
Miles Bennett
Tesla has named Intel manufacturing veteran Gary Jiang as Terafab chip-program director — the first publicly confirmed senior hire for Musk's $20 billion semiconductor plan, signaling that Tesla is trying to import the hardest part of fab-building directly from Intel's top ranks.
Who is this hire, and why does he matter?
Gary Jiang spent nearly 18 years at Intel. His most recent role: fab manager overseeing the technology transfer of Intel 18A — one of Intel's most advanced process nodes.
His track record covers construction, equipment install, product qualification, and high-volume ramp. This means → he has managed the entire journey from "empty shell" to "chips shipping at scale."
Before that, he led mass production of 14 nm and 22 nm at Intel's Ocotillo campus in Arizona. In plain terms = this is not a blueprint designer — he is someone who has physically stood up fabs and driven them to volume.
How big is Terafab supposed to be?
Terafab was jointly announced by Tesla, SpaceX, and xAI in March 2026 as a vertically integrated mega-facility — chip design, lithography, wafer fabrication, memory, advanced packaging, and test, all under one roof.
Musk's stated parameters: Intel 14A process node, initial capacity of 100,000 wafers per month, an eventual target of 1 million, and a $20 billion prototype budget.
SpaceX's S-1 filing, however, cites a much larger figure: $55 billion initial investment, with a full-phase cap of $119 billion. This means → the two numbers differ by multiples, and the project's true scale and timeline remain unclear.
What role does Intel play?
Intel formally joined Terafab in April 2026 as a manufacturing partner, contributing process technology and packaging expertise.
Jiang's move from Intel's 18A team to Tesla's Terafab mirrors that partnership. This reflects a clear talent pipeline: Tesla is recruiting from the very top tier of Intel's manufacturing organization.
Tesla has also posted other Terafab roles in Austin this month, including project managers required to have managed over $100 million in capital expenditure. In plain terms = this is not a single poach — it is a systematic effort to transplant teams from established fabs.
How large is the execution risk?
A key reality: Intel itself took decades and tens of billions of dollars to build the institutional capability to run advanced fabs — and still stumbled on leading-edge nodes in recent years.
SpaceX's S-1 filing explicitly classifies Terafab as a "very early stage" project — no binding commitments, no finalized IP rights, no defined financial terms.
This means → hiring one top executive is a necessary condition, but far from a sufficient one. The real test is whether Tesla can turn a single-point talent hire into repeatable yield, capacity, and manufacturing discipline.
How is the market reading this?
Tesla shares closed up 2% on Tuesday, extending a three-day rally, but retail-investor sentiment stayed neutral and news volume was at normal levels.
Sell-side consensus expects Tesla Q2 deliveries of roughly 406,000 vehicles, up about 6% year-on-year — the market's focus remains anchored to the core auto-delivery business.
In plain terms = Terafab is still a "signal," not an "earnings event" — the stock price reflects the car business, not the long-dated chip-factory thesis.
Content is for reference only, not financial advice.