Nvidia invests up to $3B in power firm behind Stargate

Share
Nvidia invests up to $3B in power firm behind Stargate

The AI buildout's newest bottleneck is the grid, and the industry's biggest players are moving accordingly: Nvidia is buying into the electricity developer behind Stargate, Microsoft is fast-tracking its next in-house chip, and an AI notetaker left more than 181,000 meetings exposed.

Nvidia has agreed to invest up to $3 billion in Lancium, the Texas power-infrastructure developer behind the Stargate AI campus, according to three people familiar with the deal reported by The Information. The initial $2 billion buys roughly 20 percent of the Blackstone-backed company, with another $1 billion tied to milestones like grid hookups — potentially lifting Nvidia's stake toward 30 percent and valuing Lancium's land-and-power portfolio at around $10 billion. Lancium has about 4 gigawatts of signed power in Texas, including 1.2 gigawatts feeding the OpenAI-Oracle Stargate site in Abilene and 900 megawatts for a Crusoe-built Microsoft data center, and is weighing an IPO as soon as 2027.

The deal is Nvidia's most direct move yet into electricity, the constraint now binding the AI buildout harder than silicon supply. Nvidia has also considered extending credit guarantees for a 10-gigawatt OpenAI-SoftBank project in Ohio, and its investments in private companies and infrastructure funds hit $18.6 billion in the fiscal first quarter alone. Buying into Lancium is supply-chain insurance in its purest form: the chipmaker is locking up the megawatts its customers need to run the chips it sells them, and placing itself at the center of the infrastructure stack — land, grid access, power — rather than just the compute layer.


Microsoft plans to unveil its next-generation Maia 300 AI chip this fall — possibly as soon as September — and is negotiating with TSMC to produce more than 300,000 wafers in 2027, The Information reports. Chinese outlets echoed the plans, with Microsoft signaling it will "significantly" increase AI chip production next year. Maia 300 follows the Maia 200, Microsoft's first large-scale in-house accelerator, which launched in January on TSMC's 3nm process for inference workloads. The reported timeline puts Microsoft's custom-silicon push — an explicit hedge against Nvidia pricing and supply — on a fast track, though the company has not commented.


An AI meeting-recording app left more than 181,000 meeting records exposed, including live conference IDs that let anyone walk into calls, and the security researcher who found the flaw says it went unfixed for months. The exposed Google Firebase collection spanned 181,874 records across 84,312 users and 35,003 domains, including government meetings from 23 countries; researcher BobDaHacker says she joined a live Malaysian ministry call using an exposed ID. tl;dv disputes the severity — it says the data was metadata rather than transcripts or recordings, and that it has since addressed the issue and is removing Firebase from its stack — but Dark Reading reported the exposure was still active as of August 4, six months after the initial January report. The episode is a reminder that the tools sitting inside your most sensitive conversations are only as secure as their tenant isolation.

What to watch: whether Lancium's 2027 IPO materializes, and whether Microsoft's fall event actually ships Maia 300 silicon.

Nvidia now has a financial stake in the electricity that runs its own chips — does that concentration worry you? Tell us in the comments.

Read more

Akhetonics says its all-optical CPU reaches a customer in 2026

Akhetonics says its all-optical CPU reaches a customer in 2026

Light-based computing keeps promising more than it delivers — but one Munich startup has just put a date on its bet, and the interview laying it out is doing the rounds on Hacker News this week. Akhetonics says it will deploy its first commercial machine with a major customer by the end of 2026, with several more planned for 2027. The company, founded by Michael Kissner and Leonardo Del Bino, is building a computer where data enters as light, is switched as light, and circulates through memory

The Week in AI — October 5–11, 2026

The Week in AI — October 5–11, 2026

Every big claim this week turned out to rest on fine print more interesting than the headline: revenue only the company reporting it can define, safety tests sandboxed while the product keeps the web, and a Pentagon phase-out nobody would confirm until reporters kept asking. The week's top 5 1. OpenAI's revenue was $20 billion below the numbers everyone quoted — and the gap was definitional. The Financial Times reported Thursday that OpenAI's annualized revenue runs roughly $20 billion unde

Drone strike shuts a third Yandex data center, taking YandexGPT offline

Drone strike shuts a third Yandex data center, taking YandexGPT offline

Russia's largest tech company is learning what the AI era's infrastructure war looks like from the receiving end — three data centers in four days, and with them much of the cloud layer Russian businesses run on. A Ukrainian drone strike knocked out Yandex's data center in Vladimir early Sunday morning, the third of the company's facilities hit since October 8. The site — reported at roughly 50 MW and designed for about 2,880 server racks — stopped operating completely after the attack, Yandex

Agent teams cost up to 5x more, barely score higher

Agent teams cost up to 5x more, barely score higher

The multi-agent hype train hit a benchmark this weekend — and the grid and the trucking regulators had quiet weeks of their own. Vals AI put agent teams head-to-head with single agents on its Vibe Code Bench, and the teams cost between 1.8 and 5.1 times more for almost no extra quality. The evals company ran GPT-6 Sol and Claude Opus 5.5 solo and in teams across 50 apps at two reasoning efforts; out of four comparisons, only one was statistically significant — Sol at medium effort, where the t