Etched, an AI hardware startup, has secured an additional $700 million in funding, elevating its valuation to $21 billion. This latest investment round was spearheaded by Jane Street, a prominent quantitative trading firm that has recently integrated Etched’s AI systems into its data centers.
Founded in 2022 by three Harvard dropouts, Etched has experienced a meteoric rise in valuation. In December 2025, the company was valued at $5 billion. By July 2026, it had doubled to $10.3 billion following a $300 million Series C funding round. Now, just a month later, its valuation has doubled again, marking an $11 billion increase in a remarkably short period.
Etched’s flagship product, termed “frontier inference clusters,” offers comprehensive AI solutions that encompass both hardware and software components. These systems are designed to enhance the inference process—the stage where AI models generate outputs in response to user inputs.
The inference process consists of two primary phases: prefill and decode. The prefill phase involves understanding the user’s prompt, which is computationally intensive. The decode phase focuses on generating the output, requiring substantial memory resources. To optimize these processes, Etched has developed two innovative components:
- Low-Voltage Prefill Chip: This chip operates at reduced voltage levels, mitigating heat generation and allowing for a higher density of transistors. This design enables faster processing of tokens during the prefill phase.
- Cluster-Scale Memory: This novel memory and interconnect technology facilitates rapid, low-latency communication among multiple chips, enabling them to share a common memory pool efficiently during the decode phase.
These advancements have positioned Etched as a formidable competitor in the AI hardware sector, challenging established players like Nvidia. The company’s systems are versatile, capable of running various frontier AI models, and are not limited to specific architectures.
Jane Street’s decision to lead the recent funding round was influenced by successful tests of Etched’s hardware. The firm expressed satisfaction with the early results, noting that Etched’s approach to inference meets the precision requirements for their most demanding workloads. Jane Street has since deployed Etched’s systems within its data centers.
Etched’s rapid ascent underscores the growing demand for specialized AI hardware capable of efficiently handling complex computations. As AI applications continue to proliferate across industries, the need for optimized inference solutions becomes increasingly critical. Etched’s innovative approach and swift valuation growth highlight its potential to play a significant role in shaping the future of AI infrastructure.
