Etched opens a 10-MW AI facility after hitting $10.3B valuation
Signals InboxJuly 24, 2026AI Chips
Etched has opened an 80,000-square-foot, 10-MW facility after raising $300 million at a $10.3 billion valuation. The big story is not simply another AI data center. Etched is moving from private chip demos toward producing complete inference racks, and its valuation has doubled before those systems have reached customers at scale.
We’ve raised $300M in Series C funding at a $10.3B valuation from Sequoia, Andreessen Horowitz, Jane Street, Argo, and SK Hynix.
— Etched (@Etched) July 23, 2026
Our mission is to run the world's inference. This round accelerates production of our inference clusters.
We've opened an 80,000-sqft, 10-MW facility…
Q1What actually happened?
According to Etched’s official announcement, the company raised $300 million at a $10.3 billion valuation and opened a new 10-MW lab near its San Jose headquarters. The 80,000-square-foot site will support testing, prototyping, assembly, and continuous deployment of Etched’s inference clusters.
Q2Why is the valuation important?
Because Etched was valued at about $5 billion after raising $500 million in December 2025. Seven months later, that figure has more than doubled to $10.3 billion. Investors are pricing Etched like a major AI-chip challenger before its first racks have shipped broadly.
Q3How big is a 10-MW facility?
It is meaningful for a young hardware company. Etched already operates a 2-MW data center at its San Jose office, so the new site gives it about five times more power at one location. But Etched says it wants to reach gigawatt scale, which is 100 times larger than 10 MW. This is a serious first factory step, not the final destination.
Q4Is Etched already selling chips at scale?
Not yet. Etched says its first silicon has returned from TSMC, customer testing is underway, and its first racks are due to ship this summer. It also says it has more than $1 billion in customer contracts. The tension is that demand and valuation are already huge, while mass production is only beginning.
Q5What is Etched actually competing with?
Etched is challenging the GPU model used by Nvidia and other chipmakers. Instead of building flexible processors for many workloads, it is designing complete systems around AI inference. That includes chips, memory, cooling, networking, software, and racks. The bet is that a narrower system can serve AI models faster and with less power.
Q6So what should we watch next?
Watch whether Etched ships its first racks on time, fulfills its $1 billion order book, and publishes performance that customers can verify. Raising money and opening a lab prove investor confidence. Reliable production, real deployments, and lower inference costs would prove the business.