Etched, an AI inference chip startup, raised a $300 million Series C led by Sequoia, with participat...
The AMW Read
Incremental update to a known player in the AI infrastructure segment, but valuation and round size give segment-level significance for inference chip market.
Etched, an AI inference chip startup, raised a $300 million Series C led by Sequoia, with participation from a16z and SK Hynix, at a $10.3 billion valuation. The round underscores growing demand for specialized hardware tailored to running AI models after training.
Why it matters: This fundraising sits at the intersection of the hyperscaler distribution moat and the capital-compression arc in AI infrastructure. Etched is betting that purpose-built inference silicon — not just general-purpose GPUs — will capture a growing share of the AI compute budget as model deployment scales. The involvement of SK Hynix, a memory giant, signals that the supply chain for inference-specific chips is also deepening.
Expert take: The $10.3 billion valuation for a pre-revenue or early-revenue chip startup is a signal that investors are placing large bets on the post-training compute layer. The round is large enough to fund wafer starts and tape-out costs, but the real test will be whether Etched's architecture can win against the incumbents' general-purpose accelerators. If inference demand grows faster than training, Etched could become a key player in the substrate, but it faces steep execution risk.
