Etched hits $5B valuation, books $1B in AI inference chip sales

AI semiconductor startup Etched has emerged from stealth, announcing a $5 billion post-money valuation and securing $1 billion in contract orders for its specialized inference hardware. The announcement follows the successful manufacturing of its first-generation chips at TSMC earlier this year. Etched is currently conducting customer trials of its frontier inference clusters, integrated systems that combine custom silicon, purpose-built racks, and optimized software to accelerate large language model inference. The company positions these clusters as a direct solution to the latency and cost bottlenecks currently constraining scalable AI deployment. The funding milestone includes a recently closed $500 million capital injection, bringing Etched total fundraising to $800 million. The investment round attracted a consortium of quantitative trading firms and venture capital firms, including Jane Street, Hudson River Trading, Two Sigma, VentureTech Alliance, and Ribbit Capital. Angel backing features prominent artificial intelligence researchers and industry figures, alongside billionaire investors Stanley Druckenmiller and Peter Thiel. Etched was co-founded in 2022 by Harvard dropouts Gavin Uberti and Robert Wachen, who previously served as Thiel Fellows. The startup trajectory marks a stark reversal from its early operating phase. In 2023, founders faced widespread investor skepticism regarding the necessity of specialized inference chips over general-purpose graphics processing units, with the company reportedly surviving on month-to-month liquidity. The subsequent pivot to a product-driven strategy and TSMC manufacturing completion have rapidly revitalized investor confidence. Etched rapid ascent occurs amid intensifying competition in the custom AI silicon sector. Competitors including Cerebras and Groq have recently secured substantial capital or achieved public market listings, while cloud providers and major developers such as Microsoft, Amazon, Google, and OpenAI continue advancing proprietary in-house accelerators. Etched venture into production places it directly in the race to supply cost-effective inference infrastructure, a segment identified by market analysts as the primary scaling constraint for enterprise and frontier model deployment.