Etched Raises $700M at $21B Valuation for AI Inference

In a significant validation of the burgeoning AI infrastructure sector, Etched, a San Jose-based innovator in AI inference acceleration, has successfully closed a $700 million Series D funding round. This latest capital infusion propels the company's valuation to an impressive $21 billion, more than doubling its previous market assessment from just one month prior. The round was spearheaded by Jane Street, the prominent quantitative trading firm, which has also established itself as Etched's inaugural major client.

Etched is at the forefront of developing specialized hardware and software stacks designed to dramatically enhance the speed, reduce the cost, and increase the availability of AI inference. The company's unique approach, focusing on full-stack co-design, has enabled it to achieve first-pass silicon success in under three years. Since emerging from stealth in June 2026 with a workforce exceeding 400 employees, Etched has been simultaneously advancing three distinct hardware generations, underscoring its aggressive development timeline.

The substantial Series D financing saw robust participation from a distinguished roster of investors. Key contributors include Kleiner Perkins, Sequoia Capital, Andreessen Horowitz (a16z), Tiger Global, Bain Capital Ventures, Neo, Primary, Stripes, Positive Sum, and Blackstone. This influx of capital brings Etched's cumulative funding to a formidable $1.9 billion, positioning it for significant expansion.

Jane Street's involvement extends beyond financial backing; the firm has actively integrated Etched's technology, deploying its first cluster rack into production workloads. Their early positive results offer a strong endorsement of Etched's ability to optimize demanding computational tasks, particularly for complex AI models. This real-world deployment validates Etched's core innovations, such as Low Voltage Inference (LVI) for enhanced compute density within power limits, and Cluster Scale Memory (CSM) for creating expansive shared memory pools across entire clusters.

The demand for Etched's specialized inference clusters is already evident, with the company reporting over $1 billion in secured customer contracts. These agreements span leading frontier AI organizations, private AI developers, and major cloud service providers. Etched's hardware is currently powering large-scale mixture-of-experts models and novel non-transformer architectures, demonstrating its versatility and adaptability to the rapidly evolving AI model landscape.

CEO Gavin Uberti emphasized the significance of Jane Street's production deployment, stating, "Jane Street putting this cluster into production proves what we've built. Now, we're sprinting on scaling production." The newly acquired capital will be instrumental in accelerating manufacturing capabilities and furthering hardware development to meet the escalating global need for efficient and powerful AI inference solutions. This strategic funding positions Etched to capture a substantial share of the rapidly growing AI infrastructure market, projected to reach hundreds of billions of dollars in the coming years.