AI chip startup Etched said it has raised $800 million and booked more than $1 billion in customer contracts as it prepares to ship inference systems designed to compete with Nvidia’s GPUs.

The company said its first chip was successfully manufactured by TSMC earlier this year using the N4P process. Etched is now validating its first rack scale product with customers and has started production to meet demand for its systems.

Etched calls the product a frontier inference cluster. The system combines its chips with custom racks, software and manufacturing methods built to make AI inference faster, cheaper and more power efficient.

Inference is the process that happens after a user submits a prompt to an AI model. It has become one of the largest cost centers for AI companies as more users and applications rely on large language models at scale.

The company’s latest funding includes a previously unannounced $500 million round closed in December at a $5 billion post money valuation. Investors include VentureTech Alliance, Jane Street, Hudson River Trading, Two Sigma, Ribbit Capital and Stripes, which led the round.

Etched has also attracted backing from prominent AI figures including Andrej Karpathy, Geoffrey Hinton, Fei Fei Li, Arthur Mensch and Scott Wu, along with investors such as Stanley Druckenmiller and Peter Thiel.

The announcement comes as competition intensifies across AI infrastructure. Groq recently raised $650 million to expand its inference cloud business, while OpenAI unveiled its first custom AI chip with Broadcom. Amazon, Google and Microsoft have also been building custom silicon to reduce reliance on general purpose GPUs.