AMD stories
Rising memory prices are forcing data-centre operators to size servers by workload, not total capacity, to avoid wasting budget and bandwidth.
AI cloud operators face sharply higher storage bills, with a mixed-fleet design costing almost USD $39 million less than all-flash over three years.
It aims to help businesses build AI agents that use live data, keep to strict permissions and avoid hallucinating over filings and news.
Data centre operators could cut deployment risk and speed AI build-outs with a validated blueprint for racks drawing up to 246 kW.
The tie-up is aimed at helping cloud providers cut inference costs and improve GPU efficiency as AI demand shifts beyond model training.
The deal could let enterprises serve more AI models on fewer GPUs by shifting cache data into DRAM, SSDs and remote storage.
Customers facing soaring AI compute demands will get access to one of the first cloud offerings of AMD's latest large-memory GPU and rackscale system.
Helios racks and new Instinct chips put AMD deeper into AI infrastructure as it courts major customers and challenges Nvidia.
Anthropic will gain a larger pool of AI computing power as AMD commits up to USD $5 billion and a gigawatt-scale chip rollout.
Microsoft will widen Azure's AI capacity as AMD's Helios rack system and Venice processors are added across chips, networking and software.
Businesses shifting from AI pilots to production may cut costs and tighten control as Nutanix backs private infrastructure for heavier workloads.
Engineering teams could cut AI coding bills by more than USD $1.46 million a year as Featherless offers GLM 5.2 on fixed-fee private cloud.
One million ageing Windows 10 PCs are set for replacement as Ingram Micro and Microsoft push AI-ready devices to cut security risks and cloud costs.
Eligible US universities and non-profits can now use Lightwell to patch open source flaws without overhauling systems or sharing data.
Windows 10's looming end is pushing Australian firms to upgrade PCs and security as AI use moves from pilots to daily work.
Organisations seeking tidier shared desks may adopt a keyboard-integrated PC that supports four 4K displays and local AI tasks.
Reserved capacity for new AI workloads is now open, as Vultr prepares deployments across 2027 and 2028 with AMD's latest chips.
Governments and researchers can now track land-use change, crop health and biodiversity globally without proprietary systems after Cambridge opened its dataset.
Fraud teams could cut investigation time by 90% as the new stack combines real-time scoring, case assembly and human review in milliseconds.
The game developer is using fresh capital to scale AI-driven production and live services as it faces rising costs and tougher launch economics.
