[Etched Raises $800 Million to Build Low-Latency AI Inference Clusters, Secures Over $1 Billion in Orders] AI hardware company Etched announced it is developing 'frontier inference clusters' to enhance large model inference throughput, latency, and energy efficiency through an integrated design of chips, racks, software, and manufacturing. Etched stated that its first A0 chip has completed tape-out and returned from TSMC's N4P process, and it is currently working with customers to validate the initial rack-level products to fulfill over $1 billion in inference cluster orders.