AI chip maker Etched has doubled its valuation to $10.3 billion following a $300 million Series C funding round led by Sequoia Capital. The company claims its specialized hardware significantly improves the speed and efficiency of AI inference. Investors are now tracking the firm's progress in delivering its newly manufactured chips to clients, with over $1 billion in orders already reportedly booked.
Detailed Coverage
Etched, an AI hardware startup, has secured $300 million in its latest funding round, bringing its total valuation to $10.3 billion. This rapid increase comes just seven months after the company was valued at $5 billion. The funding round was led by Sequoia Capital, with additional support from major investors including Andreessen Horowitz, SK Hynix, Jane Street, and Diffusion Capital. The influx of capital highlights strong market interest in specialized hardware designed to improve the economics and speed of AI operations.
Specialized Chips for AI Inference
Unlike general-purpose hardware, Etched focuses exclusively on AI inference—the phase where AI models process data to generate responses after being trained. The startup has developed custom silicon components to address specific bottlenecks in the 'prefill' and 'decode' stages of this process. By using a 'low-voltage inference' method, the company claims its chips can operate at higher speeds while generating less heat, which allows for greater transistor density. Additionally, Etched has introduced 'cluster scale memory' technology, which enables multiple chips to share data at low latency. This is designed to reduce the cost of running large language models, a key concern for companies deploying AI at scale.
Managing Execution and Competitive Risks
While the company has gained significant backing, it faces a complex landscape in the AI hardware sector. Etched’s business model involves providing complete hardware systems rather than just individual chips. This approach previously faced criticism from industry observers who questioned whether such specialized systems would be versatile enough to run evolving AI architectures. The company has since worked to demonstrate that its hardware can support various model types, including Mixture of Experts and non-transformer frameworks.
Another challenge for Etched is the intense competition in the AI compute space, where it must contend with established industry giants that have deeper resources and more mature supply chains. The success of the company will depend heavily on its ability to move from initial system testing to full-scale commercial delivery. With over $1 billion in orders already booked, the company is under pressure to execute its manufacturing plans efficiently and ensure its hardware meets the performance expectations of its clients.
Investors and industry watchers will be looking for updates on the actual deployment of these chips in production environments. Key future indicators will include the pace of order fulfillment, the performance of the chips in real-world data centers, and the company's ability to maintain its technological advantage as competitors continue to innovate in the semiconductor space.