GPUs

Elon Musk said xAI plans to expand its computing capacity to more than 1.2 million Nvidia GPUs by the end of the year, more than double its current count. The plan underscores the escalating race among AI companies to secure massive GPU fleets for training larger models, and highlights Nvidia's central role in supplying the global AI buildout.

China has reportedly relaxed restrictions on NVIDIA GPUs as part of a shift in its artificial intelligence strategy, according to a report in South Korean media. The move would allow Chinese firms greater access to NVIDIA's AI chips, which have been subject to US export controls and earlier Chinese procurement curbs. Details on the scope and timing of the relaxation remain limited.

China is reportedly considering allowing imports of NVIDIA GPUs as it seeks to accelerate its artificial intelligence development. The report, carried by South Korean outlet Chosun Ilbo, suggests a possible shift in Beijing's approach to securing advanced chips for its AI sector. Further details on the scope or timing of any decision have not been confirmed.

Advanced US-made GPUs, restricted by export controls aimed at China, are reportedly still flowing to buyers through an underground resale network despite tightening enforcement. The report argues that high demand, price premiums and smuggling routes make Washington's curbs largely ineffective, raising questions about whether export policy can realistically control the spread of cutting-edge AI hardware.

China is reportedly pursuing a 'hiding capabilities' strategy in artificial intelligence, built around NVIDIA GPUs. The approach suggests Chinese AI developers may be concealing the true extent of their computing power and model performance, likely to manage perceptions abroad and avoid triggering tighter export controls. The claim is circulating in international coverage of the US-China semiconductor rivalry.

Supersonic Labs has released Julia 1, an open decision model with 144.3 million parameters that is small enough to run on a standard CPU. Unlike large language models, decision models are built for making choices and taking actions rather than generating text. The low hardware requirement makes the model accessible to developers without expensive GPU infrastructure, which is drawing attention in the AI community.

Chinese regulators are considering whether to approve the sale of Nvidia's RTX Pro 5500 workstation GPUs to ByteDance and Alibaba, according to a Bloomberg report. The decision comes amid ongoing US-China tensions over advanced chip exports, with Nvidia's China business restricted by Washington's trade controls and Beijing scrutinising purchases of American semiconductor technology by its biggest tech firms.

Early benchmarks of Apple's base M5 chip show it running Resident Evil Village at 1080p with frame rates between 42 and 63 FPS, falling short of Nvidia's RTX 4060 laptop GPU. Testing suggests only the higher-tier M5 configurations can compete with high-end mobile graphics cards, tempering expectations for base-model gaming performance.

A newly shared video explores a hypothetical question gaining attention in tech circles: what would computing look like if society stopped relying on GPUs? The topic touches on how deeply graphics processors now underpin artificial intelligence, gaming and the broader economy. Discussion is focused on whether alternatives like CPUs or specialized chips could realistically fill the role GPUs currently play.

NVIDIA is again promoting the concept of GPU 'rentability', framing graphics cards as revenue-generating assets to rent out rather than devices people simply own. The framing is drawing attention and criticism in the hardware and finance communities, with commentators questioning what the term means for consumers, gamers and the company's pricing strategy.

Thieves in an unspecified location made off with a 20-ton shipment they believed contained valuable Nvidia graphics cards, only to discover the cargo was actually sand. The heist has become a viral talking point, with observers mocking the criminals' costly mix-up and questioning how the mistake happened. No word yet on arrests or the real value of what was targeted.

Developers are increasingly running AI models on Apple's Mac Mini instead of relying on Nvidia GPUs, according to a Fortune report. The shift is being attributed to the Mac Mini's lower cost and power efficiency, with Apple silicon offering competitive performance for local AI workloads. The trend highlights a challenge to Nvidia's dominance in AI hardware as smaller teams look for cheaper ways to build and test AI applications.

A speculative tech discussion is circulating asking what computing would look like if society stopped relying on GPUs, the graphics processors that now power most AI systems and gaming hardware. With GPU demand and prices surging due to the AI boom, the question of alternatives is drawing attention among developers and hardware enthusiasts.

A developer reports enabling multi-token prediction (MTP) on an RTX 3090 graphics card raised local LLM generation throughput, while questioning whether the speedup affects coding quality. The write-up, originally published on a personal blog, has drawn attention from AI and open-source software communities interested in getting more performance from consumer GPUs for running large language models locally.

Hardware company Apex Compute is developing an open-source Vulkan driver for its hardware within the Mesa graphics stack. The work means its GPUs could gain support in mainline Linux systems without proprietary drivers, a move welcomed by the open-source graphics community. Details on supported hardware and timelines have not yet been published.

Y Combinator's Startup Podcast released a new episode of its YC Paper Club series asking what would happen if computing moved away from GPUs. The discussion examines the AI industry's heavy reliance on GPU hardware and what alternatives or architectural shifts could mean for startups building on top of current AI infrastructure.

Nvidia is pressing Wall Street banks to finance AI chips using structures similar to aircraft financing, where equipment itself serves as collateral for loans. Banks, however, remain hesitant, unconvinced that rapidly depreciating and fast-obsolete GPUs offer the same durable value as planes. The debate highlights growing questions about how to fund the enormous buildout of AI datacenters and who should bear the risk if chip values fall.

Reports of market chatter suggest Amazon is considering offloading Nvidia AI chips valued at around $8 billion. If confirmed, the move would signal a shift in how the cloud giant manages its massive AI hardware commitments, and it lands amid ongoing questions about whether big tech has over-ordered GPUs. Neither company has publicly confirmed the claim, and details about timing or buyers remain unclear.

AWS has introduced the P6-B300, an instance built around Nvidia's Blackwell B300 GPUs for large-scale AI work. A QuantumBytz article argues the instance changes the real buying question for AI infrastructure: rather than raw GPU count, organizations must weigh how clusters balance compute, networking and cost. The piece is being shared in Linux and open-source tech circles, where cloud GPU pricing and cluster design are active topics among engineers planning AI deployments.

A new open-source project called Nesbox, published by Nestrilabs on GitHub, offers a fast micro virtual machine with support for sharing GPUs across workloads. It is drawing attention among developers interested in lightweight virtualization and more efficient use of accelerated hardware, a growing concern as AI and compute costs rise.

Amazon is reportedly taking a drastic financial step to fund its purchases of Nvidia chips as it ramps up spending on artificial intelligence infrastructure. The move underscores the enormous capital costs cloud providers face in the AI race, as companies compete to secure scarce Nvidia GPUs for data centers and AI services.

Amazon is seeking to offload roughly $8 billion worth of Nvidia chips, according to a report carried by Yahoo Finance. The move involves excess AI hardware capacity, and the story is drawing attention across the semiconductor and cloud computing sectors as questions grow about whether big tech has stockpiled more Nvidia GPUs than it currently needs.

Amazon is increasing the prices it charges customers to rent computing chips from its cloud service, while reportedly moving Nvidia processors off its balance sheet. The move signals tighter economics around AI cloud capacity as demand for Nvidia GPUs surges. Market Watch reported the changes, which are drawing attention because they affect costs for companies building AI services and hint at how Amazon is restructuring its GPU fleet financing.

Amazon is looking to sell roughly $8 billion worth of Nvidia chips to outside investors, according to a Financial Times report. The move would spread the cost of the company's massive AI data-center buildout, as cloud providers face enormous capital expenditure on GPUs. The news, also picked up by Seeking Alpha, is drawing attention to how tech giants are financing their AI infrastructure spending.

Austrian cooling specialist Noctua says it wants to adapt server-grade cooling capable of handling 2000W thermal loads for desktop computers. The move targets next-generation CPUs and GPUs whose power draw keeps climbing, as consumer hardware increasingly borrows thermal solutions once reserved for data centers. PC enthusiasts are watching closely to see what form the design will take.

Nvidia is in discussions with insurers about loans backed by its AI chips, according to the Financial Times. The arrangement would let buyers of Nvidia's expensive AI hardware finance purchases with the chips themselves serving as collateral. The move highlights how far Nvidia's GPUs have become a sought-after asset class as demand for AI computing infrastructure keeps climbing.

RustChain, a blockchain built around the idea that old hardware should earn as much as new machines, selects its block producers with a simple modulo operation rather than proof-of-work hashing or proof-of-stake lotteries. Commenters note the design is deliberate: a hash-power race would let modern GPUs outpace a 2003 PowerBook, defeating the project's stated goal of hardware egalitarianism.