Exclusive | A $21 Billion 'Kids in Chips' Startup Is Scooping Up Nvidia Talent
Tap title to read story
AI chip startup Etched has doubled its valuation to $21 billion following a massive $700 million funding round led by quantitative trading giant Jane Street. Co-founded in 2022 by three Harvard dropouts in their early 20s (Gavin Uberti, Chris Zhu, and Robert Wachen), San Jose-based Etched has rapidly emerged as a prominent challenger to Nvidia in the specialized artificial intelligence silicon market. Unlike general-purpose graphics processing units (GPUs) built for both training and inference, Etched designs application-specific integrated circuits (ASICs) strictly optimized for transformer-based AI inference, separating processes into low-voltage prefill stages and cluster-scale memory decode architectures. Capitalizing on its explosive growth, the company has recruited top-tier engineering talent from major legacy players—including approximately 15% of its 400-employee workforce coming directly from Nvidia, alongside veterans from Google, Broadcom, and Intel. Jane Street serves as both a primary lead investor and Etched's inaugural customer, having tested and deployed the startup's first server rack in its data center after booking over $1 billion in total customer orders.
More context
Timeline
- 2022: Etched is founded in San Jose, California, by Harvard dropouts Gavin Uberti, Chris Zhu, and Robert Wachen.
- December 2025: The startup reaches a $5 billion private valuation.
- July 2026: Etched closes a $300 million Series C funding round at a $10.3 billion valuation led by Sequoia Capital.
- August 18, 2026: The Wall Street Journal publishes an exclusive report detailing Etched's rapid ascent, recruitment of Nvidia talent, and its $21 billion valuation backed by Jane Street.
Why it matters
Immediate US impact: Technology markets and venture capital ecosystems react to the unprecedented valuation jump of an early-stage hardware company within a single month. Possible long-term US impact: Validates specialized transformer-only inference architectures, potentially threatening Nvidia's dominant market share in enterprise AI infrastructure. Most affected groups: Nvidia, enterprise data center operators, semiconductor engineers, venture capital firms, and high-frequency trading desks. Reader Priorities: Specific financial figures, the technical architecture of inference-specific ASICs, client deployment proof points, and talent acquisition trends.
Coverage notes
Left: Outlets focus on wealth concentration among elite tech founders, labor migration patterns in the semiconductor sector, and systemic capital allocation inside the artificial intelligence boom. Center: Reports emphasize verifiable financial metrics, valuation multipliers, funding round participants, and technical specifications of the startup's inference ASICs. Right: Commentary underscores market-driven innovation, free-enterprise competition challenging established monopolies, and the tangible validation provided by private corporate customers.
Primary report
The Wall Street Journal published an exclusive investigative report regarding Etched's rapid valuation surge, engineering hires, and customer deployments.
Sources
JQJO Editorial Desk · Human reviewed · AI assisted. Report a correction.
