Graphics Processing Unit (GPU) stories

Vultr has launched a USD $50m AI supercluster with 24,000 AMD GPUs at its new Ohio data centre, boosting cloud capacity for global AI development.

AWS has launched EC2 Trn3 UltraServers with new Trainium3 chips, boosting AI training speed by 4.4 times while cutting energy use and costs sharply.

Macquarie Data Centres has completed the structural phase of its AUD $350m IC3 Super West AI data centre in Sydney, opening in Q3 2026 with 47MW capacity.

Up to 84% of GPU power is wasted in multimodal AI, causing major inefficiencies and costs as firms struggle with legacy infrastructure.

AMD has launched the Instinct MI430X GPU with 432GB HBM4 memory to power AI and HPC in next-gen supercomputers across the US and Europe.

Tech giants including Amazon, Google, and Nvidia pledge GBP £31 billion to boost the UK’s AI and cloud infrastructure with new data centres and funding.

ETB Technologies sees 33% rise in GPU sales revenue, driven by AI and mixed reality markets boosting demand across multiple industries.

AI's surge demands data centres to be redesigned with liquid cooling, high-density power, and sustainability for scalable, efficient infrastructure across Europe.

Data centres face rising power and cooling demands as AI workloads push rack consumption to 300-600kW by 2030, requiring new infrastructure approaches.

Thoughtworks and Groq partner to offer faster, cost-effective AI solutions in Australia, using Groq’s LPU tech for real-time, reliable AI at scale.

AMD powers 40% of the world's top ten supercomputers, including El Capitan and Frontier, driving advances in AI and energy-efficient HPC globally.

Groq launches a 4.5MW AI inference facility in Equinix's Sydney data centre, enhancing AI compute speed and compliance across Asia-Pacific.

Pepperdata launches pepperdata.ai, an AI tool boosting GPU use to cut costs by 30%, helping firms optimise costly GPU resources in cloud and on-prem environments.

Duality Technologies partners with Google Cloud to deliver confidential AI using NVIDIA GPUs, enabling secure, scalable large language model training for enterprises.

Broadcom expands VMware Cloud Foundation with AI ReadyNodes and open networking, enhancing private cloud flexibility and easing AI integration for enterprises.

Google Cloud partners with VAST Data to deliver hybrid AI capabilities, enabling seamless enterprise AI workloads across cloud and on-premises environments.

TurinTech partners with Intel to bring offline AI code optimisation to Intel Core Ultra PCs, enhancing performance, security, and efficiency on-device.

Apple launches the 14-inch MacBook Pro with M5 chip and macOS Tahoe, boosting AI, graphics, and battery life for professionals and creatives.

Apple unveils new iPad Pro with M5 chip, OLED display, Wi-Fi 7, and iPadOS 26, boosting AI, graphics, and wireless connectivity for enhanced performance.

AMD unveils Helios, a rack-scale AI platform built to Meta's new Open Rack Wide standard, designed for high-performance AI and HPC data centres.