AI Chip Startup Etched Secures $300M Series C, Valuation Doubles to $10.3 Billion

Less than a month after unveiling its first working inference chip, Etched has closed a $300 million Series C funding round that more than doubles its valuation to $10.3 billion. The round was led by Sequoia Capital, with participation from SK Hynix, Andreessen Horowitz, Jane Street, and Diffusion — signaling strong investor confidence in the startup’s novel approach to AI inference hardware.

Founded in 2022 by Harvard dropouts Gavin Uberti and Chris Zhu, the Cupertino-based company emerged from stealth last month announcing it had developed a functional inference chip and already signed over $1 billion in customer contracts. The new financing brings Etched’s valuation to $10.3 billion, according to COO Robert Wachen in an interview with TechCrunch, more than doubling the valuation from December when it closed a $500 million round.

Etched’s chip is built around two custom-designed components that separately handle the two phases of AI inference: the compute-intensive prompt processing stage, known as “prefill,” and the memory bandwidth-intensive output generation stage, called “decode.” The company says its chip operates at a “much lower voltage than any other AI chip,” generating less heat and enabling a higher density of transistors. The hardware also uses a shared low-latency memory pool across the entire scale-up domain, and a proprietary ultra-low-latency, high-bandwidth interconnect that allows “dramatically faster” memory access between chips.

Manufactured on TSMC’s N4P process technology, the chip is now being validated with customers in rack-scale form. Etched plans to ship its first racks this summer. The startup currently employs about 400 people at its San Jose, California office, which houses a 2MW data center. According to the TechCrunch report, the company has also just opened a new 80,000 square-foot (7,432 sqm) facility in Milpitas, north of San Jose, with 10MW of capacity.

The rapid fundraising and valuation growth underscore the intense market demand for specialized AI inference chips, as hyperscalers and enterprises seek alternatives to general-purpose GPUs for running large language models. By architecting a chip that splits inference into two dedicated stages, Etched aims to deliver significant performance and power efficiency advantages — a value proposition that has already attracted major investors and billion-dollar customer commitments.