Megaspeed exec was recruited to spy on Nvidia CEO, according to US intelligence memo

techcrunch.com Nov 7, 2026

Megaspeed exec was recruited to spy on Nvidia CEO, according to US intelligence memo

Curated by Shubham · TechCapsules · Nov 7, 2026

AI-summarised brief · reviewed before publication

A U.S. intelligence memo, first disclosed by The New York Times, alleges that China’s Ministry of State Security recruited Singapore‑based Megaspeed executive Alice Huang to gather intelligence on Nvidia CEO Jensen Huang. The same memo, reported by the Financial Times, claims the two also had a romantic relationship. Nvidia responded that Jensen Huang’s interactions with Alice Huang were “not unusual” and affirmed the company does not share classified information with customers. The memo also references New Matrix, a Chinese firm cleared by the U.S. Commerce Department to purchase advanced Nvidia chips, whose listed office proved to be an empty shell, raising suspicions of a front for a blacklisted private‑equity entity. Nvidia reiterated it sells its premier technology within the United States.

💡 Why It Matters

· The allegations expose potential espionage pathways targeting a key AI hardware leader, underscoring vulnerabilities in corporate‑state interactions that could compromise cutting‑edge technology.

Capsules Nvidia

Read full article on techcrunch.com

Nvidia weighs playing the cosigner for Nscale’s AI data center in West…

Nvidia is negotiating to provide a residual‑value guarantee for billions of dollars in debt financing that AI cloud provider Nscale needs to build Anthropic’s first data center in West Virginia. The guarantee would obligate Nvidia to step in if Nscale defaults on construction payments, take temporary control of the facility and assist in finding a buyer or new tenant. Anthropic has signed a $44.6 billion contract for the initial 460‑megawatt building, which will use Nvidia’s upcoming Vera Rubin chips, and a second roughly $45 billion deal is under discussion. Nscale, aiming for a $35 billion IPO valuation, has struggled to deploy its planned 500,000 AI chips and has faced local construction setbacks. Nvidia’s support expands beyond typical lease and chip‑purchase assistance, deepening its financial exposure to a neocloud that has already received $1 billion of convertible financing from the chipmaker.

💡 Nvidia’s guarantee ties its balance sheet directly to a high‑risk data‑center build, signaling that chip makers are becoming de‑facto financiers to keep AI infrastructure pipelines moving.

techcrunch.com · Oct 29

Voice AI in vogue: Nvidia releases new open transcription model

Nvidia unveiled Nemotron 3.5 Transcribe, a free, open‑weight model that transcribes audio in up to 40 languages and translates 31 languages to and from English. The 30‑billion‑parameter model handles recordings up to an hour long, distinguishes up to eight speakers, and claims top performance on the Hugging Face Open ASR Leaderboard. Nvidia also released Granary v2, an expanded open‑source speech dataset, and promotes integration with its NeMo Speech toolkit for industry‑specific jargon. The model runs on a single chip, from Jetson AGX Thor to Blackwell data‑center GPUs.

💡 By offering a high‑performance, multilingual transcription tool that scales from edge devices to data‑center GPUs, Nvidia expands the practical reach of AI‑powered voice applications in business and healthcare, sectors that demand reliable, industry‑specific speech recognition.

nextplatform.com · Oct 27

Nvidia Simulates Its Way Into Google’s Magic State Cultivation Party

On October 15, Google Quantum AI unveiled the first hardware implementation of magic‑state cultivation on its Willow superconducting chip, achieving a 40‑fold error reduction and 99.99 % logical fidelity for T‑states. A week later Nvidia demonstrated a 104‑qubit simulation of the same technique on Jupiter, Europe’s inaugural exascale system powered by GH200 GPUs, claiming a 90 % qubit‑usage reduction versus traditional methods. Magic‑state cultivation, introduced by Google in 2024, refines noisy states into high‑fidelity resources without discarding large volumes, easing the overhead that previously consumed up to 95 % of quantum capacity. Both firms presented the advances as pivotal steps toward scalable, fault‑tolerant quantum computers, with Nvidia emphasizing the role of large‑scale simulation in designing next‑generation processors for future commercial quantum applications worldwide.

💡 By slashing the qubit overhead needed for magic‑state production, the technique clears a major bottleneck, bringing practical fault‑tolerant quantum processors within nearer reach.

techcrunch.com · Oct 27

Nvidia backs Indian startup helping companies run AI models with fewer GPUs

Nvidia has led a $14 million Series B round for Indian AI‑inference startup Simplismart, valuing the four‑year‑old company at roughly $100 million. The Bengaluru firm offers software that automatically tailors AI model deployment to existing GPU infrastructure, allowing customers to balance speed, cost and performance and, in some cases, run workloads with fewer GPUs. Clients such as InVideo, HeyGen and cloud provider Yotta have reported significant reductions in GPU serving expenses, with InVideo cutting costs by 56 percent. Simplismart’s platform can be used on public clouds, on‑premises data centers or via its own service, and it also enables cloud providers to build AI inference offerings. Nvidia will collaborate on technical integration, including its Dynamo inference stack, while promoting the startup to potential buyers globally.

💡 Making inference cheaper widens the market for Nvidia’s GPUs, turning efficiency gains into broader hardware adoption.

nextplatform.com · Oct 26

Nvidia Simulates Its Way Into Google’s Magic State Cultivation Party

Nvidia and Google announced breakthroughs in magic‑state cultivation, a technique that reduces the overhead of generating high‑fidelity non‑Clifford gates for universal quantum computers. Google’s Quantum AI team demonstrated the method on its Willow superconducting chip, achieving a 40‑fold error reduction and 99.99 % logical fidelity for T‑states. Days later Nvidia ran a 104‑qubit simulation of the same process on Jupiter, Europe’s first exascale machine powered by GH200 GPUs, claiming a 90 % qubit savings versus conventional distillation. Both companies say the results move the field closer to fault‑tolerant quantum computing by proving the approach works in hardware and by providing a scalable simulation platform for future processor design. The collaboration highlights how industry‑scale supercomputing can accelerate quantum algorithm validation and hardware planning today.

💡 Slashing qubit demand makes practical quantum advantage attainable sooner, sidestepping the need for massive hardware expansions.

techcrunch.com · Oct 24

Hottest toy for Christmas? Nvidia CEO plays with adorable Microduck robots, 20,000…

A video of Nvidia CEO Jensen Huang playing with Microduck robots during his GTC Berlin keynote went viral, prompting a surge in pre‑orders that topped 20,000 units by the following day, according to Hugging Face co‑founder Clem Delangue. The open‑source Microducks, produced by Pollen Robotics—acquired by Hugging Face in 2025 and now part of Nvidia after its September purchase—respond to voice commands, walk, sit, fall asleep, and perform somersaults. They were trained in Nvidia’s RTX Spark AI stack, with simulations teaching the basic tricks shipped with each unit. Owners can program new behaviors and share them via the community platform. Delivery is slated for the holiday season, positioning the toys as a potential top‑selling robot for Christmas.

💡 The viral demo showcases Nvidia’s AI ecosystem delivering consumer‑ready robotics, turning a developer‑focused platform into a mainstream holiday product.