Majorana 2 halves Microsoft's timeline to a scalable machine and reports a 20-second parity lifetime. The preprint measured one observable on one nanowire, unreviewed; critics say that is not yet a demonstrated topological qubit.
By April 2026, Amazon's random-graph fabric had become the default for most new AWS datacenters. The efficiency claims behind it are still Amazon's own, with no independent benchmark yet.
Britain is funding homegrown silicon for machines that, for now, run on an American vendor's chips. Whether that buys sovereignty or quietly rebrands dependence is the question the spending leaves open.
The GPU supply chain has the industry's attention. But the constraint that increasingly decides when an AI factory energizes is no longer the chip. It is power delivery, and the licensed electricians who commission it.
Meta's first end-to-end account of running a 150 MW, 83,000-GB200 cluster - when power is the ceiling, the cluster, not the chip, is what you optimize.
NVIDIA is bringing receipts to ICRA 2026 - eight accepted papers it says show the sim-to-real gap closing on real hardware. More and more, a robot policy is the output of large-scale GPU simulation, with the heavy compute done long before a gripper touches an object.
Multi-year mandatory funding for a mix of government-owned, contractor-operated, and commercial-surge compute... reversing the July 2025 White House AI Action Plan that told DoW to lean on hyperscalers.
A federal judge halted NSF's handoff of the NCAR-Wyoming Supercomputing Center. The problem wasn't legality; it was that NSF's process was 'arbitrary and capricious,' and the expertise running the machine can't be quickly replaced.
Finland's new 33.9 PF Roihu supercomputer triples national capacity and shares its hydroelectric Kajaani site with EuroHPC's LUMI - a two-tier sovereign-plus-EuroHPC campus under one operator.
NextSilicon's Maverick-2 dataflow accelerator has met Sandia's Vanguard system-acceptance requirements on HPCG, LAMMPS, and SPARTA. The harder question comes this fall, when Sandia decides whether to move Spectra toward more demanding, mission-like ASC supercomputing workloads.
Supercomputing News is built from first principles for both humans and AI agents. Agents discover content through a machine-readable manifest, pay per request in USDC on Base, re-read anything they have already bought for free with a wallet proof, fetch clean Markdown instead of scraped HTML, and verify every article against an on-publication Ed25519 signature. No accounts. No API keys. No quotas.
GET /llms.txt402 · X-PAYMENTSIGN-IN-WITH-Xscn-pubkey.jsonAccept: text/markdown
IBM told investors it plans to spend more than $10 billion on quantum over five years. Annualized, that is roughly twice the U.S. federal QIS R&D budget, but the figure is a blended forward-looking plan, not a booked balance-sheet item.
Commerce signed nine LOIs worth $2.013 billion to anchor a U.S. quantum supercomputing supply chain. Two foundry models funded in parallel, federal equity in every recipient.
An arXiv preprint shows single-molecule ODMR in a designer carbene with 2.2 ms coherence at 4.5 K. Two days later, NVision announced a $55 million Abbott-led round and a photonic molecular-qubit architecture still demonstrated at just one emitter.
At Computex 2026, Wiwynn and eight ecosystem partners showed a full optical scale-up rack built around compute-side optics. It is a useful moment to separate two technologies the supercomputing industry keeps filing under one acronym.
Normal Computing's CN101 is in characterization. Extropic has a prototype platform, an MIT-co-authored arXiv preprint, and an ETH Zurich hackathon in June. After two years as a manifesto, thermodynamic computing is producing the kind of artifacts readers can evaluate.
Hardware has reached orbit and SpaceX has filed for a million-satellite constellation. Thermal physics, launch cadence, and bandwidth still push gigawatt orbital AI to post-2030, at best.