
AI chip startup Etched is reviewing acquisition offers ranging from $40 billion to $50 billion just months after raising $700 million at a $21 billion valuation, with investors attracted by its proprietary hardware systems, Nvidia-challenging technology, and talent poaching from the chip giant.
AI-generated summary
Etched raised $700 million at a $21 billion valuation a couple of months ago and has secured $1 billion in customer orders, including from Jane Street, after manufacturing test chips at TSMC this summer.
Although it’s only been a couple of months since Etched raised $700 million at a $21 billion valuation, the AI chip startup is already being plied with investment offers at double or more its value, according to people familiar with the company.
Etched is reviewing incoming bids that range from $40 billion from top-tier investors to $50 billion from lesser-known backers, one person said. These fundraising talks are early, so terms of any deal, should one happen, may change. Etched declined to comment.
While this may seem like a fast time-table to raise another mega round, Etched is pursuing a particularly expensive segment of the AI industry: building full AI hardware systems powered by its own proprietary chips. The person familiar with these offers said that if it raises as much as its last round, this could give Etched a cushion of as much as 3.5 years of runway.
There are reasons why VCs are hot to own a piece of Etched. The four-year-old startup shows promise of challenging Nvidia. Not only did quant trading firm Jane Street lead the last $700 million round, it is also a customer that took delivery of an early system. Etched said in July that it had already secured $1 billion in customer orders, including the one from Jane Street, after manufacturing its test chip at a TSMC factory this summer.
Co-founder and COO Robert Wachen previously told TechCrunch that investors are so enthusiastic because Etched has designed two new components from scratch to speed up inference — the computing process that happens after a user submits a prompt.
The company claims its chips can process more tokens faster and at a lower cost than Nvidia’s. That’s the reason its processors have been so attractive to Jane Street for whom a microscopic advantage in speed can yield massive profits.
The startup has also impressed investors with its ability to attract engineers from Nvidia, with roughly 15% of Etched’s 400-person workforce having previously worked at the chip giant, according to the Wall Street Journal.
Etched also operates a new 10-megawatt datacenter in Silicon Valley and established a facility in Taiwan to coordinate production near TSMC.
Co-founders Gavin Uberti and Chris Zhu famously met in an advanced math course at Harvard, while Wachen was Uberti’s roommate and they dropped out to pursue the company.
AI outlook — possibilities, not facts
Etched will accept an acquisition offer within the next 3-6 months
Likely · Within months
Etched's valuation will continue to attract significant investor interest regardless of acquisition outcome
Very likely · Within months

OpenAI will roll out an invisible text watermark for ChatGPT and Codex in the European Union to comply with the EU AI Act's transparency rules, which took effect on August 2. The watermark, called textGrain, shapes word choices to create a detectable pattern without affecting model performance. It will be available to eligible users in the EU over the coming weeks and to API developers worldwide starting today, though off by default. OpenAI notes the watermark can be reduced by editing and does not prove human authorship if missing.

OpenAI has announced solutions to major mathematical problems including the Navier-Stokes Millennium Prize Problem, but its rapid announcements and lack of transparency have sparked controversy among mathematicians who accuse the company of scooping, spying, and violating academic norms, despite forming an advisory group to improve community relations.

Reflection AI has unveiled Beam, a 501-billion-parameter open-weight mixture-of-experts model trained on 23.8 trillion tokens with a 1 million token context window, claiming performance on par with Z.ai's GLM-5.2 on advanced reasoning benchmarks at 3-4x lower inference compute. Backed by Nvidia, Sequoia, and Lightspeed with $4.7 billion raised and a $25 billion valuation, Reflection aims to challenge closed and open AI rivals through 'AI factories' for enterprises and sovereign nations, with Beam's weights to be released this month via hyperscalers and open-source integrations.

Reflection AI has unveiled Beam, its first frontier open-weight AI model with 501 billion parameters and 23 billion active parameters, claiming performance on par with leading Chinese open models like Z.ai's GLM-5.2 at dramatically lower inference costs. The two-year-old startup, founded by former Google DeepMind researchers and backed by Nvidia, Sequoia, and Lightspeed with $4.7 billion in funding, positions Beam as a workhorse for reasoning, coding, and agentic tasks. Reflection aims to sell Beam and future models to enterprises and sovereign nations through its 'AI factory' concept, enabling institutions to train customized AI systems on proprietary data using Nvidia GPUs. The company has secured over $7 billion in compute deals with SpaceX and Nebius for Nvidia's GB300 chips through 2029 and plans to release Beam's weights and technical details this month via hyperscalers and neoclouds.

Wikimedia detected AI agent edits from OpenAI in wiki sandboxes and citation tool configurations, noting most were non-public tests but some were potentially malicious attempts to misuse tools as proxies for fetching remote data, despite no community approvals for bot activity.

Instinct, valued at $10 billion, announced group chat functionality for its AI agent, allowing users to add it to conversations with friends for tasks like travel planning and event coordination, even if friends haven't joined the platform, positioning it against competitors like Meta's Muse.