Etched announced on Tuesday that it has raised an additional $700 million at a $21 billion valuation, led by Jane Street. The valuation jump is remarkable, especially in the AI sector. Etched was valued at $5 billion in December and raised a $300 million Series C at a $10.3 billion valuation in July. Now, investors have more than doubled its valuation to $21 billion in just a month. Etched delivers its AI technology as full systems called 'frontier inference clusters.'
Co-founder and COO Robert Wachen explained that investors are enthusiastic due to Etched's new components designed to accelerate inference. These components speed up the computing process after a user submits a prompt. 'Inference is built in two stages,' Wachen said, 'prefill and decode.' In the prefill phase, the system must understand the prompt, including context. In the decode phase, the system generates output tokens, which are the actual answers users see. Etched created a prefill chip that operates at low voltage, enabling more transistors without typical heat issues. It can process more tokens faster. The company also developed a new type of memory and an interconnect for the decode process, called cluster-scale memory. 'It allows many chips to connect together and use a shared memory pool at a very, very fast, low latency,' Wachen said. The result is higher speeds and lower costs.
Etched is still facing the perception that its chips are custom-designed for a specific model. This was its original intention but is no longer the case. Etched's systems can run any frontier model. In the blog post announcing the new round, investment firm Jane Street said, 'We tested the chip and are pleased with the early results. Etched’s unique approach to inference delivers the precision we will need to support our most demanding workloads. We’re excited to now have our own rack running in our datacenter.' Other investors include Kleiner Perkins, Sequoia Capital, Andreessen Horowitz, Peter Thiel, Tiger Global, Bain Capital Ventures, Neo, Stripes, Primary, Positive Sum, Diffusion, Argo, and Blackstone.
Source: techcrunch