# Nvidia rival Etched raises $800M with backing from Jane Street and a TSMC-linked fund > Source: > Published: 2026-06-30 15:47:55+00:00 #### TL;DR *AI chip startup Etched has raised $800M backed by Jane Street and a TSMC-linked fund, with one billion dollars in sales contracts and summer shipments planned.* The Harvard-dropout-founded chip startup says it has signed one billion dollars in sales contracts and will start shipping this summer *AI chip startup Etched has raised $800M backed by Jane Street and a TSMC-linked fund, with one billion dollars in sales contracts and summer shipments planned.* AI chip startup Etched has raised $800 million and revealed that its backers include trading firm Jane Street and VentureTech Alliance, a venture firm with a strategic partnership with TSMC. The company, which designs chips specifically for running AI models rather than training them, says it has signed one billion dollars in sales contracts and plans to start shipping to customers this summer. The bulk of the funding came from a $500 million round led by Stripes that closed in December, valuing Etched at five billion dollars. That round included Peter Thiel, Positive Sum, Ribbit Capital, Hudson River Trading, and Two Sigma. Jane Street separately led a previously unannounced round and has invested more than $100 million in total, [according to Bloomberg.](https://www.bloomberg.com/news/articles/2026-06-30/ai-chip-startup-etched-says-jane-street-tsmc-linked-vc-invested?srnd=phx-technology) The investor list reads like a who’s who of AI. Geoffrey Hinton, who won a Nobel Prize for foundational work in modern AI, is an investor. So are computer vision pioneer Fei-Fei Li and hedge fund manager Stanley Druckenmiller. Founded in 2022 by Harvard dropouts Gavin Uberti and Robert Wachen, Etched has been quiet for roughly two years while building its product. The company’s chip, called Sohu, is designed to run transformer models by embedding the architecture directly into silicon rather than relying on general-purpose GPUs. Working with TSMC, Etched developed what it calls low-voltage inference, running chips at lower voltage to prevent overheating and squeeze more performance from the hardware. Etched also designed its entire server rack, including circuit boards, cooling plates, and networking connections, rather than just the chip itself. No other chip startup has done this, Wachen told Bloomberg. The company has 400 employees, more than half based near its San Jose headquarters. The inference chip market is attracting enormous capital as the industry shifts from training models to running them at scale. [Nvidia paid Groq $20 billion](https://thenextweb.com/news/groq-650-million-raise-nvidia-20-billion-inference-cloud) in December for a licensing deal that took most of its engineers. Google announced in April that a version of its own AI chips will focus on inference. [London-based Fractile raised $220 million](https://thenextweb.com/news/fractile-220m-inference-chip) for inference chips that put compute and memory on the same die. The race to build purpose-built silicon for inference, rather than repurposing training GPUs, is now one of the most capital-intensive bets in the semiconductor industry. Whether Etched can deliver on one billion dollars in contracts depends on whether its chips perform as promised under production workloads, a question no startup in this space has fully answered yet. “*If you have compute now, people will buy it,*” said Positive Sum CEO Patrick O’Shaughnessy. The startup’s bet is that being early with a full-stack rack, not just a chip, gives it an edge that matters more than benchmark numbers on a spec sheet. Get the most important tech news in your inbox each week.