AI chip startup Etched, founded by three Harvard dropouts in 2022, has closed a $300 million Series C at a $10.3 billion valuation — double its December valuation of $5 billion. The round was led by Sequoia, with participation from Andreessen Horowitz, SK Hynix, Jane Street, Diffusion Capital, and earlier investors. The company claims this is the highest valuation ever for a Sequoia-led Series C. Last month, Etched announced it had successfully manufactured its homegrown chips, that its first full systems were being tested by clients, and that it had already booked $1 billion in orders.

Etched designs chips specifically for transformer-based AI models (the architecture behind ChatGPT and Claude), an idea that was initially considered risky. The company is still fighting the perception that its products — sold as full systems, not just chips — can only run specific LLMs. Co-founder and COO Robert Wachen explains that the systems can run any AI model, including Mixture of Experts models like DeepSeek and Qwen, as well as non-transformer designs like Mamba (based on state-space models). Google is reportedly pursuing a similar concept with its Frozen v2 chip for Gemini.

Etched’s key innovation is in speeding up inference, which happens in two stages: prefill and decode. Prefill understands the prompt and is compute-intensive; decode generates output tokens and requires massive memory bandwidth. Etched created a prefill chip that runs at “dramatically” lower voltage than other AI chips — a technique called low-voltage inference — which generates less heat and allows more transistors. For decode, the company built a new memory and interconnect technology called cluster scale memory, which lets many chips share a memory pool with very low latency. The result, Etched claims, is high speed at lower cost.

Despite the progress, the founders have faced persistent skepticism. Much of it comes from limited access to the hardware; only investors and early customers have seen it. That’s how Etched landed its famous backers — by showing private demos in its office. Wachen names Andrej Karpathy, Noam Brown, and Geoffrey Hinton as people who “actually tried the hardware and are very excited about it.”

The road has been long and difficult. The founders dropped out of Harvard without knowing how to raise money or hire. Wachen recalls sleeping on a friend’s floor with a towel as a blanket. Early servers were set up in an employee’s garage, and his wife had to reboot the machine when needed. Today, Etched has 400 employees and operates a 2-megawatt data center. “We’re running tokens in our lab today, working with some of the largest AI companies in the world,” Wachen says. He now has a mattress and multiple pillows. “I think, when you really think something’s possible, and you just work at it for a long time, you can do it.”