Welcome back to SavvyMonk, your daily dose of AI and tech news that actually matters.

This weekend, Elon Musk stood inside a decommissioned power plant in Austin and announced the most ambitious semiconductor project any private company has ever attempted.

Let's get into it.

The AI economy is booming, and smart entrepreneurs are already profiting. Subscribe to Mindstream and get instant access to 200+ proven strategies to monetize AI tools like ChatGPT, Midjourney, and more. From content creation to automation services, discover actionable ways to build your AI-powered income. No coding required, just practical strategies that work.

Subscribe to Get Your Free Guide

On March 21, Musk took the stage at the historic Seaholm Power Plant in downtown Austin, Texas, and officially unveiled Terafab. It's a joint venture between Tesla, SpaceX, and xAI (which SpaceX acquired in an all-stock deal in February 2026 for a combined valuation of $1.25 trillion).

The estimated price tag: $20 to $25 billion, though Morgan Stanley analysts think the real number could land between $35 and $45 billion once scaling costs are included.

The facility will be built on the North Campus of Giga Texas, right next to Tesla's existing Austin headquarters. And the plan is to bring every stage of chipmaking under one roof: design, lithography, fabrication, memory production, advanced packaging, and testing. No semiconductor company on Earth currently does all of that in a single location.

Musk called it “the most epic chip-building exercise in history by far.” Texas Governor Greg Abbott was in the audience.

Why Musk Says He Needs It

The short version: his companies are running out of chips.

At Tesla's Q4 2025 earnings call, Musk warned investors that external chip capacity from TSMC, Samsung, and Micron would hit a ceiling within three to four years. At the Terafab event, he put it bluntly: “We either build the Terafab or we don't have the chips, and we need the chips, so we build the Terafab.”

He estimated that total global AI compute output sits at roughly 20 gigawatts per year. That figure, he said, represents about 2% of what his companies will eventually need. Read that again. He's saying Tesla, SpaceX, and xAI alone will need roughly 50 times the entire world's current AI chip output.

Two Chips, Two Worlds

Terafab is designed to produce two distinct chip families. The first is an edge-inference processor for Tesla's Full Self-Driving system, the Cybercab robotaxi, and the Optimus humanoid robot line.

Tesla's fifth-generation AI chip, AI5, is among the first products targeted, with small-batch production reportedly aimed at late 2026 and volume production in 2027.

The second is D3, a high-power, radiation-hardened chip built specifically for space. It's designed to run hotter than terrestrial chips to minimize radiator mass on satellites. This chip would power SpaceX's planned orbital data center constellation. In January, SpaceX filed an FCC application to launch up to one million data center satellites into low Earth orbit.

The Space Angle

Here's where the pitch goes full sci-fi. Musk said 80% of Terafab's compute output would ultimately be directed toward space-based AI satellites. His argument: solar irradiance in orbit is roughly five times greater than at Earth's surface, and heat rejection in the vacuum of space makes thermal scaling viable. He projected 100 to 200 gigawatts per year of terrestrial chip output, with the remainder (up to a terawatt) going to space.

His claim is that within two to three years, running AI workloads in orbit will be cheaper than running them on the ground. It's a bold prediction from a man who also said we'd have fully autonomous Teslas by 2020.

The Skeptic's Case

There's plenty of reason to pump the brakes.

Tesla, SpaceX, and xAI have never fabricated a single chip. Terafab's full-capacity target of one million wafer starts per month would represent roughly 70% of TSMC's entire current global output. TSMC spent decades and hundreds of billions of dollars building that capability.

The $20-25 billion cost estimate isn't even included in Tesla's 2026 capital expenditure plan yet, which already exceeds $20 billion. Tesla's 2025 revenue declined 3%, and its automotive revenue dropped 10%. The company's free cash flow last year was $6.2 billion on $8.5 billion in capex. Now it's proposing to more than double that spending while simultaneously funding a multi-decade semiconductor project.

And then there's the talent question. Since the SpaceX-xAI merger, 9 of xAI's 11 original co-founders have departed. Musk publicly acknowledged that xAI “was not built right the first time around.”

Musk gave no timeline for when Terafab would begin producing chips or hit its stated targets. Bloomberg noted that Musk has no background in semiconductor production and a well-documented history of overpromising on timelines. Battery Day in 2020 promised a revolution in cell manufacturing. Five and a half years later, the 4680 program is widely considered a disappointment.

But here's the Thing

People have counted Musk out before. Repeatedly. Reusable rockets were supposed to be impossible. Mass-market electric vehicles were supposed to be a pipe dream. SpaceX now dominates commercial launch, and Tesla remains the world's most valuable automaker by market cap.

The core logic behind Terafab is sound. AI chip demand is growing faster than existing foundries can scale. Vertical integration, if achieved, would give Musk's companies a structural advantage no competitor could easily replicate. And if SpaceX's IPO (expected as early as this summer at a potential $1.5-1.75 trillion valuation) generates the capital needed, the funding question gets a lot easier to answer.

The real question isn't whether Musk can envision this. He clearly can. It's whether he can execute it. Chip fabrication isn't software. It isn't even rocket science. It's arguably harder.

The Bottom Line

Terafab is either the most important infrastructure project of the AI era or the most expensive PowerPoint presentation since Battery Day.

The need for AI chip scaling is real. The vision of space-based compute is genuinely interesting. But building a world-class fab from scratch, with no prior experience, while simultaneously running three companies and advising a government, is a bet that would make even the most aggressive venture capitalist blink. Watch what gets built, not what gets announced.

Category: Strategic Analysis

“I'm evaluating a major capital expenditure decision for [Company Name]. The project costs [Amount], has no precedent in our industry, and requires capabilities we don't currently have in-house. List the top 10 risks I should present to the board, ranked by likelihood and potential impact. For each risk, suggest one mitigation strategy.”

Project Hail Mary opened at number one this weekend with $80.6 million. A movie about one guy on a spaceship trying to save the world. Musk wants to put a million satellites in orbit to run AI. Sometimes reality doesn't wait for fiction to finish its opening weekend.

Hit reply, I read every response.

See you tomorrow.

— Vivek

P.S. Know someone who follows the intersection of AI, chips, and space? Forward this their way. They can subscribe at https://savvymonk.beehiiv.com/