GoKawiil
Tech News
Mark Weatherford, a security executive at Nvidia, says chief information security officers need to actively help business leaders evaluate the benefits and risks of deploying AI systems. He argues that the teams and leaders who choose to deploy AI should also be accountable for managing the risks that come with it.
SpaceX disclosed a binding commitment for about $40 billion in financing to acquire AI chips and data center hardware, calling it the largest committed GPU financing deal to date. An Apollo-managed entity will purchase the equipment and lease it back to SpaceX for roughly five years, with the deal expected to close by year-end and leases beginning in early 2027 as equipment arrives.
Meta is launching roughly $35 billion in stock and convertible preferred securities this week, with a further $60 billion share-sale program planned once new nonvoting Class C shares begin trading in early 2027. The company says the fundraising will help finance its AI infrastructure expansion, including four superclusters each exceeding one gigawatt of capacity, as annual capital spending is now forecast at $135 billion to $145 billion. Nvidia, AMD and Broadcom have reportedly expressed interest in buying up to $5 billion of the shares combined.
At GTC Berlin, Nvidia CEO Jensen Huang said running the company's AI agent software on each GPU costs roughly $2,000 per year, calling software licensing an increasingly major part of Nvidia's business. The figure, reported by The Register, is lower than Nvidia's existing AI Enterprise suite pricing of $4,500 per GPU annually, or $3,600 under a five-year deal, and it remains unclear how the two pricing schemes relate or when the new figure might take effect.
Elon Musk announced on X that SpaceX has closed $40 billion in debt financing to buy GPUs for AI computing, with funds expected to arrive in early 2027. He said the financing is mostly callable debt and that SpaceX qualifies as investment grade due to revenue from Starlink and its rocket launch business. Bloomberg had earlier reported the company was negotiating with banks including Goldman Sachs and investors like Apollo Global Management to raise the funds for Nvidia chips.
At GTC Berlin on October 22, Nvidia CEO Jensen Huang said licensing the company's software to run AI agents on its 'AI factories' would cost around $2,000 per GPU annually. He did not specify how many GPUs would be covered, which products the fee includes, or whether this represents a standard or final price.
Nvidia has entered a preliminary partnership with German energy firm PreussenElektra to evaluate whether its decommissioned nuclear power plant sites could be converted into AI data centers. The assessment, announced October 21st, will examine grid connections, land availability and existing industrial infrastructure at the former plants to see if they can support high-performance computing facilities.
Block, Jack Dorsey's company, released a new voice model called Dessa from its research lab, trained entirely on human speech rather than text and runs on a single NVIDIA GPU. The model is full-duplex, meaning it can speak and listen simultaneously, and reportedly produces natural-sounding speech with laughs, pauses, and emotional cadence built directly into the audio rather than driven by text or voice tags.
Tom's Hardware Premium published its weekly roundup covering TP-Link's efforts to satisfy US regulators amid scrutiny over its China ties, including a letter of intent with Key Tronic to assemble Wi-Fi 8 routers in Mississippi and nearly $1 billion pledged toward US R&D and manufacturing. The roundup also details testing of Nvidia's new RTX Spark-based Microsoft Surface Laptop Ultra for gaming performance, plus an examination of whether costly audiophile-grade headsets are worth it for gamers.
Nvidia has built an extensive financing role in the AI buildout, including roughly $36 billion in off-balance-sheet cloud agreement backstops and guarantees worth up to $105 billion tied to a proposed OpenAI data center in Ohio. Shares of neocloud partners CoreWeave, Nebius and IREN have each fallen at least 30% this month as credit markets tighten, putting Nvidia's financing commitments under scrutiny.
Apple's new Apple TV 4K 2026 has launched priced at $299, more than double the $129 starting price of the previous generation, matching the recently hiked Nvidia Shield TV Pro. Android Authority highlights five alternative streaming boxes, including the Google TV Streamer 4K, which despite its own price rise from $99 to $149 still costs about half as much while offering 4GB of RAM and strong performance.
A new open-source Linux kernel module called GPU-unlock removes Nvidia's artificial cap on FP16/BF16 instruction issue rates for GeForce RTX 3000-series cards based on the GA102, GA103, GA104 and GA106 chips, bringing them up to the rate normally reserved for professional cards. It works by writing specific speed-selector registers and performing a function-level reset before the stock Nvidia driver loads, avoiding the need for manual driver patching after each update. Community-reported benchmarks on cards like the RTX 3090 Ti, 3090, 3080 and 3070 show BF16 GEMM throughput gains ranging roughly from 25% to over 80%.