AI chip startup Etched hits $10.3B valuation with top investor backing

Three Harvard dropouts took a big idea and turned it into a company now valued at over $10 billion. Etched, an AI chip startup founded in 2022, has closed a $300 million Series C funding round at a $10.3 billion valuation, according to co-founder and COO Robert Wachen. The round was led by Sequoia, with participation from Andreessen Horowitz, SK Hynix, Jane Street, and Diffusion Capital, alongside earlier backers including Peter Thiel, Andrej Karpathy, Dylan Field, and Amjad Masad.

This marks a rapid ascent. Just seven months ago, in December, Etched was valued at $5 billion after raising $500 million. The company claims this is the highest valuation ever for a Sequoia-led Series C. Last month, Etched announced it had successfully manufactured its own chips, that its first full systems were undergoing client testing, and that it had already secured $1 billion in orders.

Etched launched when the concept of building a chip specifically for transformer-based AI models,the architecture behind systems like ChatGPT and Claude,was considered far-fetched. The company still contends with the perception that its products, sold as complete systems rather than standalone chips, are limited to running only specific large language models.

That is not accurate, Wachen explains. The systems can run any AI model, including Mixture of Experts models like DeepSeek and Qwen, which split tasks across specialized sub-models, as well as non-transformer designs like Mamba, built on a state-space model architecture. Interestingly, the idea of etching parts of a specific AI model directly into silicon is no longer seen as radical. Google is reportedly pursuing a similar approach with its Frozen v2 chip for Gemini.

Etched’s core innovation lies in two custom components designed to accelerate inference,the computing process that occurs after a user submits a prompt. “Inference is built in two stages,” Wachen says, “prefill and decode.” The prefill phase involves understanding the prompt and its context, requiring heavy mathematical computation. The decode phase generates the output tokens, demanding less computation but massive memory bandwidth.

For the prefill stage, Etched created a chip that operates “dramatically” faster “by running at a much lower voltage than any other AI chip. We call this low-voltage inference.” Lower voltage reduces heat, allowing the chip to pack in more transistors. For decode, Etched developed a new memory and interconnect technology called cluster scale memory, which enables many chips to connect and share a memory pool with very fast, low latency. The result, the company promises, is high speed at lower cost.

Because Etched launched before most of the tech world,outside Nvidia,understood AI’s specialized compute needs, founders Gavin Uberti, Robert Wachen, and Chris Zhu have faced persistent skepticism, even after announcing that their first silicon batch was successfully manufactured by TSMC.

Much of that doubt stems from limited access. So far, only investors and early customers have seen the systems. In fact, that is how Etched landed its famous backers,by showing them private demos in its office. “Andrej Karpathy from Anthropic, Noam Brown from OpenAI, Geoffrey Hinton, as well as all the investors in the funding round,these are all people who actually tried the hardware and are very excited about it,” Wachen says.

Still, the road has been long and difficult, and mass production and delivery of rack systems remain ahead. The trio famously dropped out of Harvard to launch Etched, with no prior experience raising capital or hiring. “We had no idea how hard it was going to be,” Wachen says. “I think we still have to be humbled by what it will take to actually get to scale.”

Wachen recalls landing in the Bay Area after telling his parents he was leaving school for a startup, with no office or apartment arranged. He slept on the floor of a friend’s unfurnished house. “I remember staying in my friend’s house that they were about to sell, using a towel as a blanket,” he laughs. The founders eventually set up servers for chip-design tools in an early employee’s garage. “Every time it needed to be rebooted, he would call his wife, and she would go and hit the reboot button.”

Today, Etched has 400 people working in an office 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 also has a blanket now, and a mattress “and a pillow even. Multiple pillows,” he jokes.

More importantly, he and his co-founders never let the doubters stop them. “It’s come a long way. It’s a very, very different world. But I think, when you really think something’s possible, and you just work at it for a long time, you can do it.”