An AI chip startup just tripled its valuation in six months. Here's who's paying and why.
Etched has raised $700 million at a $21 billion valuation, led by a Wall Street trading giant that actually tested the hardware first. The numbers are striking. The speed of the rise is even more so.
Key points
- Etched raised $700 million in a funding round announced Tuesday, valuing the AI chip startup at $21 billion.
- Jane Street, a quantitative trading firm known for using algorithms and data to make financial bets, led the round after testing Etched's hardware in its own data centre.
- Etched was valued at $10.3 billion in July 2025 and $5 billion in December 2024, meaning the company's valuation has roughly doubled in roughly one month.
- Etched's chips are designed specifically to speed up inference, the process that happens when an AI model generates an answer to your question.
- Backers include Kleiner Perkins, Sequoia Capital, Andreessen Horowitz, Blackstone and Peter Thiel, among others.
The valuation clock on Etched is moving fast. The AI chip startup was worth $5 billion in December 2024. It hit $10.3 billion in July 2025 with a $300 million Series C. Now, barely a month later, investors have handed it another $700 million at a $21 billion valuation. That is nearly $11 billion added in four weeks.
The lead investor is Jane Street, a quantitative trading firm that uses sophisticated computer models to trade financial assets at enormous scale. Jane Street is not a typical Silicon Valley venture fund. It put money in only after testing the hardware, then installed a rack of Etched chips in its own data centre.
"We tested the chip and are pleased with the early results," Jane Street wrote in the announcement. "Etched's unique approach to inference delivers the precision we will need to support our most demanding workloads."
What does Etched's hardware actually do?
Etched builds specialised chips and the full server systems that house them, which it calls frontier inference clusters. The key word is inference, the computing work an AI model does the moment you hit send on a prompt.
When a large language model, the technology inside chatbots like ChatGPT and Claude, answers a question, it runs through two stages. First comes the prefill phase, where the system reads and processes your entire prompt. Then comes the decode phase, where it writes out the answer, word by word.
Etched built a separate custom chip for each stage. The prefill chip runs at low voltage, so it can pack in more transistors, the tiny switches that do computing work, without overheating. More transistors, processed faster, means it can handle more text at once. For the decode stage, Etched built a new type of shared memory system it calls cluster-scale memory, letting many chips pool their memory as if it were one giant resource.
Co-founder and COO Robert Wachen told TechCrunch that the combination delivers higher processing speeds at lower cost per query.
What's worth watching here?
Valuations like this can feel abstract, but the money trail tells a real story. Jane Street's decision to test before investing matters. A firm of that size running Etched chips in production is a meaningful commercial signal, not just a bet on a pitch deck.
That said, survivorship bias is real in AI hardware. For every chip startup that lands a marquee customer, several others raise at sky-high valuations and quietly disappear. Etched is also still shaking off an early reputation for building chips locked to a single AI model. The company now says its systems run any frontier model, a flexibility that matters if the AI landscape shifts.
If Etched delivers on cost and speed, it could be worth watching for any business that pays large inference bills to providers like OpenAI or Google.
One honest takeaway: If your business spends meaningful money on AI tools, the cost of running AI models is a real line item and it is being competed down. More chip options, not fewer, tend to lower prices over time.
