Etched valuation doubles to $21 billion in new funding
AI hardware startup Etched has raised $700 million at a $21 billion valuation, doubling its worth in just a month as it deploys specialized systems to accelerate frontier model inference.
Etched announced on Tuesday that it has secured $700 million in a new funding round led by quantitative trading firm Jane Street. This investment values the startup at $21 billion, marking a massive leap from its $10.3 billion valuation during a $300 million Series C round in July, and its $5 billion valuation in December. The rapid valuation step-up reflects intense investor appetite for alternative AI hardware architectures as companies scramble to find more efficient ways to run massive models.
The startup delivers its hardware as full systems called frontier inference clusters, positioning them as direct competitors to Nvidia's AI factories. According to co-founder and COO Robert Wachen, the systems target the two distinct stages of the inference process: prefill and decode. For the compute-heavy prefill phase, Etched designed a low-voltage chip that packs in more transistors without generating excessive heat, allowing it to process prompt contexts faster. For the memory-intensive decode phase, the company developed a custom interconnect and memory architecture called cluster-scale memory, which links multiple chips to a shared, low-latency memory pool.
For AI practitioners, these architectural choices translate to faster token generation speeds and reduced operational costs. Crucially, Etched has moved past its original concept of hardcoding a single specific model onto its silicon. The company's current systems are flexible enough to run any frontier model, eliminating the risk of hardware obsolescence when new models are released. Jane Street has already deployed an Etched rack in its datacenter, stating that the chip delivers the "precision we will need to support our most demanding workloads."
The funding round also drew participation from a prominent roster of venture firms and tech figures, including Kleiner Perkins, Sequoia Capital, Andreessen Horowitz, Peter Thiel, Tiger Global, Bain Capital Ventures, Neo, Stripes, Primary, Positive Sum, Diffusion, Argo, and Blackstone.
This is our own summary of reporting by TechCrunch AI
