Jane Street lidera la financiación de 700 millones de dólares de Etched y, luego, despliega el chip por sí misma

The Cryptonomist · 2026-08-19T10:08:13.000Z
Etched ha logrado uno de los saltos de valoración más rápidos en la carrera por el hardware de IA de este año. La startup, que construye sistemas de computación especializados para la inferencia de IA, ha cerrado una ronda de financiación de 700 millones de dólares que valora la empresa en 21.000 millones de dólares, una cifra aún más impactante si se considera que Jane Street, el inversor líder de la ronda, también es el cliente de pago más reciente de Etched. La ronda de financiación de Etched de 700 millones de dólares marca el segundo gran salto de valoración para la compañía en apenas un mes, y viene acompañada de una historia de implementación en el mundo real, no solo de entusiasmo por parte de los inversores.
Etched has pulled off one of the fastest valuation jumps in the AI hardware race this year. The startup, which builds specialized computing systems for AI inference, has closed a $700 million funding round that values the company at $21 billion — a figure made even more striking by the fact that Jane Street, the round’s lead investor, is also Etched’s newest paying customer. The Etched $700 million funding round marks the second major valuation jump for the company in barely a month, and it comes with a real-world deployment story attached, not just investor enthusiasm.
Key takeaways
Etched raised $700 million at a $21 billion valuation, led by quantitative trading firm Jane Street.
Jane Street became Etched’s first customer and received the company’s first shipped inference cluster rack last month.
Etched exited stealth mode in June with a working chip and more than 400 staff, and says it achieved first-pass silicon success within three years of its seed funding.
The company has secured more than $1 billion in customer contracts from AI firms and cloud providers.
Jane Street separately agreed in April to spend roughly $6 billion on CoreWeave’s AI cloud platform, which will incorporate Nvidia’s Vera Rubin technology.
Etched Secures $700 Million Funding Led by Jane Street
The headline number here is the speed. Etched was valued at $5 billion back in December, then jumped to $10.3 billion in July on the back of a $300 million Series C round. Just weeks later, the Etched $700 million funding round pushed that figure to $21 billion — nearly doubling the company’s worth in roughly a month, according to TechCrunch.
Jane Street led the new round, with backing from a long list of familiar Silicon Valley names: Sequoia Capital, Kleiner Perkins, Andreessen Horowitz, Bain Capital Ventures, Stripes, Positive Sum, Tiger Global, Neo, Primary, Blackstone, along with investor Peter Thiel and firms Diffusion and Argo. What sets this round apart from a typical AI hardware raise is that the lead investor didn’t just write a check — it also became Etched’s first customer, putting its own money behind hardware it had already tested and deployed.
Beyond the Jane Street relationship, Etched says it has locked in more than $1 billion in customer contracts spanning both public and private AI firms as well as cloud providers. That pipeline suggests the company isn’t relying on a single anchor client to justify its valuation, even if Jane Street’s endorsement is doing much of the heavy lifting in the current narrative.
Jane Street Becomes First Customer and Deploys Etched’s AI Hardware
Jane Street isn’t just backing Etched financially — it’s already running the startup’s hardware inside its own data center. The quantitative trading firm received Etched’s first shipped rack last month and is now using it to support live workloads, marking a rare case where an investor-turned-customer can point to actual production use rather than a pilot promise.
Early Technology Deployment and Customer Experience
Jane Street’s own account of the rollout carries real weight given its reputation for rigorous technical vetting. “We tested the chip and are pleased with the early results,” the firm said. “Etched’s unique approach to inference delivers the precision we will need to support our most demanding workloads. We’re excited to now have our own rack running in our data centre.”
Why does this matter? Quant trading firms live and die by latency and precision, so a positive endorsement from Jane Street carries more signal than a typical enterprise pilot. If Etched’s clusters can hold up under the kind of demanding, low-latency workloads that quantitative trading requires, that’s a meaningful proof point for other potential customers watching from the sidelines.
Etched’s Rapid Product Development and Technical Innovations
Etched’s pitch to investors rests as much on speed of execution as on the technology itself. The company says it achieved first-pass silicon success within three years of securing its seed funding — a milestone that, in chip design, often takes far longer and multiple costly revisions to reach.
Development Milestones and Staff Growth
Etched came out of stealth mode in June with a working chip already in hand and a headcount that had grown past 400 employees, shortly before closing its Series C round. The company is now simultaneously developing three generations of hardware at once, according to the firm — an unusually aggressive pace for a startup still ramping up production for its first wave of customers.
Co-founder and CEO Gavin Uberti framed the Jane Street deployment as validation of that pace: “We’ve felt the urgency to get our hardware into customers’ hands and run real workloads since day one. Jane Street putting this cluster into production is proof of what we’ve built. Now, we’re sprinting on scaling production for the rest of our customers.”
Innovative AI Hardware Technologies
Etched describes its products as “frontier inference clusters” — full systems rather than standalone chips, a framing that echoes how Nvidia talks about its own “AI factories.” According to co-founder and COO Robert Wachen, the company built two components from scratch to speed up the two core stages of inference: the compute-heavy “prefill” phase, where a system processes and understands a prompt, and the memory-intensive “decode” phase, where it generates the actual output.
Etched calls its prefill technology Low Voltage Inference, a design that runs at low voltage to pack in more transistors without the heat issues that plague other high-end AI chips, allowing it to process more tokens faster. For the decode stage, the company built what it calls Cluster Scale Memory — a new memory and interconnect approach that Wachen said “allows many chips to connect together and use a shared memory pool at a very, very fast, low latency.” Etched says the combination of the two technologies delivers higher speeds at lower costs, and that its clusters are already running large mixture-of-experts models as well as non-transformer designs. Notably, the company has also moved past its early positioning as a chipmaker focused on a single frontier model — its systems are now designed to run any frontier model.
Strategic Industry Partnerships and Infrastructure Expansion
Jane Street’s bet on Etched is part of a broader infrastructure buildout, not an isolated hardware purchase. The firm had already agreed in April to spend approximately $6 billion to use CoreWeave’s AI cloud platform, a separate deal that predates its investment in Etched.
Under an expanded partnership, CoreWeave will supply Jane Street with computing infrastructure across multiple data center locations, incorporating Nvidia’s Vera Rubin technology. Taken together with the Etched investment, the moves show a quant trading firm building out AI infrastructure on two fronts at once — leaning on an established cloud giant for scale while betting early on a startup chip designer for specialized inference performance.
For the wider AI hardware market, Etched’s jump to a $21 billion valuation in a matter of weeks signals just how much capital is chasing companies that can prove real deployment, not just benchmarks on paper. Whether that pace of investment holds as Etched scales production beyond its first customer remains the next test to watch.
FAQ
Who led Etched’s recent $700 million funding round?
Jane Street led the funding round and also became Etched’s first customer.
When did Etched ship its first AI inference cluster rack to Jane Street?
Etched shipped its first rack to Jane Street last month.
What are some key technological achievements of Etched so far?
Etched achieved first-pass silicon success within three years of seed funding, exited stealth in June with a working chip and over 400 staff, and is developing three generations of AI hardware at the same time.
How is Jane Street utilizing Etched’s technology?
Jane Street is actively deploying Etched’s AI inference clusters within its own operations and has said it is pleased with the early test results.
Article produced with the assistance of artificial intelligence and reviewed by the editorial team.