Etched, the AI chip startup building what it calls frontier inference clusters, raised $700 million in a Series D round on August 18, 2026, at a $21 billion valuation — roughly doubling its worth in under a month after closing a $300M Series C at $10.3B in late July.
The round was led by Jane Street, the quantitative trading giant, in an unusual sequence: the firm stress-tested Etched's hardware for its most demanding workloads, purchased a rack for its own datacenter, and then decided to lead the funding round. We tested the chip and are pleased with the early results, Jane Street said in the announcement.
Founded in 2022 by three Harvard dropouts — CEO Gavin Uberti, COO Robert Wachen, and Chris Zhu — Etched designs custom silicon optimized for AI inference rather than training. The company's systems split the inference pipeline into two stages: a compute-intensive prefill phase and a memory-intensive decode phase, each handled by purpose-built components. The result, Etched claims, is higher speeds at lower cost than general-purpose GPUs.
The investor roster reads like a cross-section of AI capital: Kleiner Perkins, Sequoia, Andreessen Horowitz, Peter Thiel, Tiger Global, Bain Capital Ventures, and Blackstone all participated. The company has already booked $1 billion in orders and began shipping systems manufactured with TSMC in June 2026.
Etched says the new funding will go toward building new factories, expanding global supply chains, and developing fleet software as it scales toward what it calls Gigawatt-scale deployment. The company also noted it now supports Mixture of Experts architectures like DeepSeek and Qwen, as well as non-transformer designs like Mamba — countering earlier perceptions that its chips were locked to a single model architecture.
