From the archive
AI hardware provides the computing power, memory and networks used to train and run artificial intelligence models. Follow news on GPUs, AI accelerators, data centres, chip design and energy demand. Read how changes in infrastructure affect performance, availability and cost, and distinguish measured results from vendor claims about what the next generation of computing will deliver.
Lead story
Researchers have identified a shared geometric structure in independently trained vision and language models that could allow capabilities to be transferred in a single forward pass. The important point is not another marginal benchmark win, but a possible way to make models work together without retraining them from…
Read the report & discussion ↗Watch Desk
Alibaba is rolling out a new AI accelerator that it describes as China’s most powerful, with the chip intended to support a planned 20GW expansion of data-centre capacity by 2032. The important point is not just another chip announcement. Alibaba is tying domestic silicon to a very large…
Watch Desk
Google has introduced GKE Pod snapshots, a new Kubernetes feature that saves a running workload’s state, including CPU and GPU memory, so it can be restored on demand. The practical prize is faster AI inference startup, with Google claiming latency reductions of up to 89% and a route…
Watch Desk
TinyTorch is a new free, open-source curriculum from PyTorch that takes learners from basic tensors to transformer models by having them build a working machine-learning framework in pure Python. It runs on an ordinary laptop without a GPU, making the machinery behind modern AI considerably less mysterious and…
Watch Desk
California has signed seven bills tightening oversight of its fast-growing data-centre industry, with new requirements covering electricity costs, water use and local control. The package matters because AI infrastructure is no longer just a Silicon Valley server-room question. It is becoming a public bill, a planning dispute and,…
NVIDIA Watch
Nvidia is pushing deeper into server CPUs as AI agents create demand for the computing work around each model call. The important shift is that the agent boom may expand the market for processors, memory and virtual machines, while Nvidia tries to sell the whole rack rather than…
Watch Desk
Samsung is putting up to $100 million behind Kairos Power’s plan to build a 50-megawatt demonstration reactor in Tennessee, with Google lined up to receive nuclear power for AI data centres by 2035. The deal matters because the AI infrastructure race is now reaching into the power system…
Watch Desk
AI inference is no longer one neat burst of computation. A new SemiAnalysis article explains how modern models divide serving into distinct stages, each placing different demands on chips, memory, networking and software. The practical consequence is that faster AI may depend as much on moving data efficiently…
Microsoft AI Watch
Microsoft is urging capital-markets firms to stop bolting AI onto isolated tasks and redesign their investment operations around governed data, reusable workflows and human accountability. The argument matters because it moves the AI conversation from clever document tools to the plumbing of decisions, permissions and institutional memory.
Watch Desk
A US congressman is calling for a national strategy to govern the rapid expansion of AI data centres, arguing that a state-by-state approach can leave local communities paying for the infrastructure. The proposal matters because it would move the argument beyond approving individual facilities and towards setting national…
Watch Desk
A national opposition bus tour stopped in Berkeley County, West Virginia, on Friday, putting a proposed 548-acre data-centre campus at the centre of a wider argument about who pays for AI infrastructure and who gets a say. The visit matters because the debate is moving from abstract concerns…
Microsoft AI Watch
Microsoft has launched India South Central, its fourth cloud region in India, in Hyderabad, adding AI-capable infrastructure, three-zone resilience and what it describes as zero-water cooling. The practical point is straightforward: organisations in India and the wider region now have another local option for running cloud and AI…
Watch Desk
The European Commission has proposed that data centres in Europe with capacity above 500 kW disclose how efficiently they use energy and water, using an EU-designed labelling system. It is a modest-looking policy move with a very large target: the infrastructure now being built to power cloud computing…
AMD Watch
AMD has introduced ROCm Hyperloom, an open-source multi-agent harness that profiles, optimises and validates AI inference workloads on AMD Instinct GPUs. AMD says it delivered validated throughput improvements ranging from 1.35x to 7.31x across supported serving frameworks, without human intervention. That makes this more than another software cupboard…
Watch Desk
Two Newark advocacy groups are asking for a one-year moratorium on new and expanded AI data centres, arguing that the city should study their environmental and public-health impact before approving more projects. The campaign matters because Newark’s council has backed a moratorium resolution, but that resolution does not…
Watch Desk
Tim Dettmers is launching dlab Open Source Week with tools for running AI locally, building more efficient autonomous research systems and cutting the token bill with a technique called CliffCompaction. His broader argument is that the next wave of AI research should not belong only to frontier labs…
NVIDIA Watch
A report covered by Euronews estimates that operating NVIDIA chips sold since 2022 could have produced between four and 21 million tonnes of carbon dioxide in 2025, depending on electricity sources and usage. The important point is not that one number has magically become the number, but that…
Hugging Face Watch
Hugging Face has published a detailed preview of tokenizers v1, a major rewrite aimed at stopping text preparation from leaving increasingly fast AI systems waiting around. The project reports large gains in decoding throughput, 76% scaling across eight native threads on an Apple M4 Max, and identical token…
DeepSeek Watch
DeepSeek is planning to use more domestic chips to train its AI models, with Huawei expected to begin supplying training processors as early as the fourth quarter, according to The Information. The move would deepen the partnership between two Chinese technology companies trying to reduce their exposure to…
ASML Watch
ASML is working with TSMC, Nvidia, Intel, Samsung and SK Hynix on High NA EUV tools that could print larger AI processors, with a pilot line targeted for 2031. The important point is that the next chip race may depend on making the canvas bigger, not merely drawing…
