In this week's issue:

  • Intel just answered the biggest question about Terafab — and it changes everything about how Musk's $25B chip factory actually gets built
  • Austin produced the steepest rent decline of any major US city, and the skyline comparison to San Francisco is genuinely embarrassing for the coastal model
  • A UT engineering student built the most comprehensive AV tracking platform in Austin using a traffic cam pipeline — 7 million rows of data, in his spare time
  • 22% of Tesla's Austin robotaxi fleet is now running unsupervised, and a new Texas regulation could shut it all down by May 28
  • A self-driving car killed a duck in Mueller and the AV company has opened a formal investigation

Let's get into it.

Quick Top Stories

Top Stories

Feature #1

Intel Joins Terafab — What It Actually Means

We covered Terafab when it was announced in March: Musk's $20-25B semiconductor project at Giga Texas, targeting 1 terawatt per year of AI compute for Tesla's robotaxis, Optimus robots, and SpaceX's orbital data centers. The obvious skeptical question was: none of these companies have ever built a chip fab. How does a carmaker, a rocket company, and an AI lab manufacture 2nm-class silicon at scale? On April 7, Intel officially answered that question. Intel's corporate X account posted that it is "proud to join the Terafab project with SpaceX, xAI, and Tesla to help refactor silicon fab technology," citing its ability to "design, fabricate, and package ultra-high-performance chips at scale." Musk confirmed the same day. Intel CEO Lip-Bu Tan, who met with Musk over the prior weekend to seal the deal, called it "exactly what is needed in semiconductor manufacturing today." Intel stock moved up roughly 3% on the news.

Now the more interesting question: what exactly did Intel agree to? The honest answer is that the full structure is not yet public — no formal press release, no SEC filing, the entire announcement happened via X posts. But the weight of analytical evidence points toward an Intel Foundry services arrangement rather than a fully greenfield, build-from-scratch fab at Giga Texas. TechPowerUp's analysis suggests Intel's existing fab facilities in Oregon and Arizona would likely "become part of the network needed for the Terafab project, while the Terafab facility itself conducts custom work guided by Intel." Electrek framed it as Tesla co-anchoring an Intel Foundry expansion. This distinction matters: if the Giga Texas North Campus becomes an assembly, packaging, and custom-work hub drawing on Intel's established wafer fabs — rather than a complete vertical semiconductor factory built from nothing — the timeline and cost structure look different from Musk's original "most epic chip-building effort ever" framing. The 1 TW/year compute target would correspond to roughly 160,000 wafers per month using Intel's 14A node (2nm-class), per analyst estimates. That is not nothing. That is enormous.

What Intel gets out of this is equally significant. The foundry business has been Intel's most troubled division — cut capacity just as AI demand exploded, losing ground to TSMC, struggling to sign anchor customers. Terafab gives Intel two anchor customers simultaneously: Tesla and SpaceX/xAI. That is a foundry turnaround thesis in two clients. The US government currently owns an 8.9% stake in Intel following a direct investment in August 2025, which adds a national-security dimension the parties haven't emphasized but that hangs over the whole arrangement. One more item worth noting: while Musk's companies are announcing partnerships at Giga Texas North Campus, Travis County separately withheld some Tesla incentive payments over "insufficient reporting" — despite Tesla reportedly exceeding its job creation and investment targets by wide margins. Tesla apparently promised 5,000 jobs and delivered something like 15,000. Travis County's compliance machinery penalized them anyway. The bureaucracy finds a way.

Sources: Intel on X, Elon Musk on X, BSCN thread, Tom's Hardware, CNET, TechCrunch, Austin Business Journal on Travis County

Upcoming Events

Feature #2

Austin Keeps Building — The Skyline That Tells the Story

San Francisco passed Proposition M in 1986, capping annual office development at 475,000 square feet and making new towers structurally impossible. Austin never did anything like that. The result, four decades later, is a comparison that makes the coastal model look genuinely indefensible. Austin's CBD median proposed building height went from 203 feet in 2010 to over 500 feet by 2024. Two-thirds of all Austin buildings over 300 feet were built after 2014. The Waterline — at 1,025 feet, now the tallest building in Texas — is under construction in a city that, 25 years ago, counted the state capitol as its tallest structure. San Francisco's per capita GDP is approximately $324,000 versus Austin's $100,000, which sounds like a win for the scarcity model until you understand the mechanism: SF achieved it by filtering out everyone who couldn't afford the constrained supply, not by creating more total value. Austin's skyline is a record of capacity. SF's is a record of gatekeeping that drove up the price of what already existed for 40 years while nothing new got built.

The data underneath Austin's skyline transformation is equally clear. Through a specific, deliberate deregulatory policy stack — targeted upzoning near jobs and universities, parking minimum elimination, ADU legalization on nearly every lot, and streamlined permitting — Austin produced the steepest rent decline of any large US city. Rents are down 7% year-over-year with approximately 50,000 new apartment units absorbed into the market. This is not an economic accident. It is the direct, measurable result of policy choices made deliberately over a decade. The Metropolitan Abundance Project put it plainly: Austin built its way to affordability while other cities regulated themselves into crisis. No rent control, no subsidy programs, no government-managed allocation schemes — just supply, allowed to happen.

And Austin is still adding tools. The city's new Accessory Commercial Use (ACU) policy legalizes customer-facing businesses under 200 square feet — coffee stands, artist studios, cottage food operations — directly in residential zones without requiring full commercial rezoning. It is fine-grained deregulation that compounds the broader housing reforms: the same city that eliminated parking minimums is now allowing a coffee window in someone's front yard without triggering a year-long permitting nightmare. The cumulative effect is a city building vertically, building densely, and now building commercially at the neighborhood scale simultaneously. Austin did not arrive at this moment by accident. It arrived by consistently choosing to build over choosing to restrict — and the skyline, the rent data, and the ACU policy are all telling exactly the same story.

Sources: Aakash Gupta on Austin vs SF skylines, Metropolitan Abundance Project on rent decline, Jonathan Berk on ACU policy

Weird Austin

The Exit

One Thing

Big issue today — Intel, Terafab, a skyline that proves the build-everything model works, and a duck. If this is the kind of coverage that belongs in your inbox:

  • Forward this to one person building something in Austin right now
  • Reply and tell me what I got right or wrong — I read every one
  • Share on X if anything here was genuinely useful

Thanks for being here. Austin is where this is all happening.

Forwarded to you? Subscribe at atxdaily.news and get it every day.

Keep Reading