Terrafab: Tesla + SpaceX, AI rocket science

Tesla and SpaceX Just Started Building the Biggest Factory on Earth - YouTube

by  The Electric Viking

This video reports on the rapid start of construction at Terafab, a massive joint semiconductor manufacturing facility being built by Tesla and SpaceX in Grimes County, Texas.

Key Highlights:

• Unprecedented Scale: The facility is planned to span over 100 million square feet, which would make it the largest building on Earth by a significant margin—reportedly 10 times larger than Tesla's existing Gigafactory in Texas (1:51-2:15).
• Purpose: The mega-factory is designed to mass-produce AI chips necessary for Tesla's self-driving vehicles, Cybercabs, Optimus robots, and SpaceX technology (3:13-3:31).
• Investment and Progress:
  • Construction has moved rapidly since its official announcement in early August 2026, with foundation work, grading, and footings already underway (0:00-0:30, 1:22-1:49).
  • The first phase is budgeted at approximately $16.3 billion USD, with total project costs potentially exceeding $100 billion (2:46-3:03).
  • Intel is reported to be a partner, contributing advanced chip manufacturing expertise (3:08).
• Timeline: Musk is targeting initial mass production by 2028, a timeline that would be record-breaking for a chip fabrication plant of this size (3:55-4:03).

Tesla + SpaceX $16.8 Billion Terafab Will Be 10X Bigger Than TSMC's Largest Fab - YouTube 

This video discusses a major joint initiative between Tesla and SpaceX to construct a massive semiconductor manufacturing campus known as Terafab in Grimes County, Texas (0:16). Here are the key highlights:

• Project Scope and Scale: The project aims to be the largest chip manufacturing site in the world, spanning over 100 million square feet—reportedly 10 times larger than TSMC's largest facility (1:40-1:48). The first phase of construction involves a $16.8 billion commitment (0:22, 3:56).
• Strategic Rationale: Both companies have identified compute power, rather than just batteries or software, as the primary constraint on their future growth. The facility is intended to supply chips for Tesla's Optimus robots, Cybercabs, and Cybervans, as well as future SpaceX orbital data centers (0:47-1:10).
• Facility Design: The plant is planned as a vertically integrated hub that consolidates logic, memory, packaging, and testing operations under one roof, targeting a capacity of 1 million wafer starts per month (2:14-2:21).
• Site Sustainability: The facility will be located at the former Gibbons Creek coal-fired power plant site (closed in 2018). It is designed to draw water from the existing reservoir, avoiding the use of local groundwater and positioning the project as a symbolic part of the energy transition (3:10-3:41).
• Caveats: While the project has a significant footprint, current filings describe Terafab as a general framework without finalized binding IP agreements between the two companies (4:03-4:12). Tesla continues to maintain other supply arrangements with industry partners like Samsung and TSMC (4:21-4:33).
  
  
  

How SpaceX Streamlined the Raptor Engine - by Brian Potter

3D printing "rocket science"

The article details how SpaceX simplified the external design of its Raptor rocket engine from version 1 to version 3 while increasing thrust by roughly 35%.

Here are the key takeaways from the engine's evolution:

• Engine Architecture: The Raptor operates as a full-flow staged combustion (FFSC) engine. This complex design routes all propellant through preburners to drive the engine's pumps, improving efficiency and theoretical reliability. The fundamental architecture has not changed between versions.
• Component Reduction: The messy tangle of wires on Raptor 1 was largely due to diagnostic sensors needed for development, which were eliminated in later versions. SpaceX also removed the main chamber spark igniters.
• Manufacturing Updates: To reduce mass and prevent leaks, many bolted flange connections were replaced with welded joints.
• Internalized Plumbing: Relying heavily on 3D metal printing, SpaceX moved much of the engine's external piping and cooling channels directly into the internal structure of the components.
• Mass Reduction: By internalizing the plumbing, SpaceX was able to eliminate the bulky external heat shields and fire suppression systems, drastically reducing the overall mass of the engine hardware.
  
  

While the Raptor 3 appears vastly simpler on the outside, this clean aesthetic was achieved by shifting the engineering complexity to the inside of the engine.

Google Spanner Omni DB; Paxos replication protocol

famous Google cloud db can now run on premise, too, as a single exe file!

Spanner Omni | Google Cloud

Spanner Omni is a self-managed database that utilizes Paxos-based synchronous replication and a software-only TrueTime API to deliver global ACID transactions and high availability across distributed servers.

What is Paxos-Based Synchronous Replication?

Paxos is a foundational distributed consensus algorithm designed to help a cluster of independent computers agree on a single value or state, even if network partitions occur or individual nodes fail. Paxos-based synchronous replication applies this algorithm to database design. It requires that before any write transaction is finalized and acknowledged to the client, a strict majority (quorum) of the replicas must formally vote to accept and persist the update.

How It Works: Roles and Phases

The algorithm functions by dividing responsibilities into three distinct roles (though a single database node typically performs all three simultaneously):

• Proposers: Receive write requests from the application and propose them to the rest of the cluster.
• Acceptors: Evaluate and vote on the proposals. A response from an acceptor signifies its agreement to store that specific value.
• Learners: Execute the final agreed-upon value and apply it to their local storage once consensus is finalized.
  
  

To safely reach this consensus, the protocol generally executes in stages:

1. Prepare: A proposer establishes leadership (using a unique sequence or ballot) and checks if the network has already committed to a previous value.
2. Accept: The proposer submits the new data value for the acceptors to vote on.
3. Commit: Once a majority of acceptors agree, the value is officially chosen, and the learners commit the change to the database log.
   
   

The Trade-offs of Synchronous Consensus

• Strong Consistency: Because a quorum must agree before a write is completed, the replication is inherently synchronous. This guarantees strict consistency (linearizability)—ensuring that any subsequent read operation instantly reflects the latest write across the distributed system.
• Fault Tolerance: The system remains highly available as long as a majority of nodes are online. In a standard 5-node cluster, the database can seamlessly survive the failure of 2 nodes without data loss or downtime.
• Increased Latency: The primary drawback is speed. Gathering votes across a network (especially a geographically distributed one) takes time, resulting in higher write latency compared to simple asynchronous leader-follower replication.
  
  

Real-World Example: Google Spanner

As highlighted in the Software Engineering Daily episode, this distributed consensus protocol is the absolute core of Google's globally scaled database architecture. Spanner utilizes Multi-Paxos, an optimized variant that uses a stable leader to efficiently handle a continuous log of transactions rather than running the full voting phase for isolated values. This architecture ensures global synchronization, extreme scale, and ironclad data durability.