AI Infrastructure

MoE racks redefine frontier AI strategy, costs, and deployment choices

MoE racks redefine frontier AI strategy, costs, and deployment choices

Melissa Palmer
December 3, 2025

NVIDIA bets MoE frontier models on GB200 NVL72 rack-scale GPU fabric NVIDIA is positioning mixture of experts (MoE) architectures as...

digi power x pivots from cypto with Modular Tier III B200 data centers

digi power x pivots from cypto with Modular Tier III B200 data centers

Melissa Palmer
December 3, 2025

Digi Power X pivots from crypto to AI with modular Tier III B200-powered data centers and a neocloud GPU service....

Liquid-cooled AI servers show a practical path to lower energy

Liquid-cooled AI servers show a practical path to lower energy

Melissa Palmer
December 2, 2025

Lenovo is pushing its Neptune direct water-cooling across more servers to cut data center power by about 40 percent versus...

Sovereign‑ready AI factories by hpe and nvidia reshape data center planning

Sovereign‑ready AI factories by hpe and nvidia reshape data center planning

Melissa Palmer
December 2, 2025

HPE and NVIDIA Push “AI Factory” Blueprint for Sovereign, GPU‑Dense Data Centers HPE and NVIDIA expanded their partnership to deliver...

VMware’s Answer to Private AI

VMware’s Answer to Private AI

Melissa Palmer
November 29, 2025

VMware Private AI Foundation with NVIDIA is VMware’s flagship approach to bringing generative AI and LLMs directly into the enterprise data...

NVIDIA Locks In Quantum-GPU Beachhead With NVQLink Across Global Supercomputing Centers

NVIDIA Locks In Quantum-GPU Beachhead With NVQLink Across Global Supercomputing Centers

Melissa Palmer
November 29, 2025

NVIDIA announced that more than a dozen leading supercomputing and national research centers in the US, Europe, Middle East, and...

What AI Really Looks Like Inside the Enterprise

What AI Really Looks Like Inside the Enterprise

Melissa Palmer
November 29, 2025

I will admit faster than anyone that I live in the nuts and bolts of AI data centers and infrastructure....

AI memory and storage advances reshape how future data centers are planned

AI memory and storage advances reshape how future data centers are planned

Melissa Palmer
November 28, 2025

SK hynix is pushing hard on AI memory bandwidth and capacity with 12-layer HBM4 (2,048 I/O, ~40% better power efficiency)...

Penguin solutions Planning Blackwell-era AI clusters as integrated stacks, not DIY builds

Penguin solutions Planning Blackwell-era AI clusters as integrated stacks, not DIY builds

Melissa Palmer
November 25, 2025

Penguin is pre-validating OriginAI clusters built around NVIDIA DGX B300 (Blackwell) and RTX PRO 6000 GPUs, targeting both large training...

AI infrastructure becoming a national-scale industrial play in US

AI infrastructure becoming a national-scale industrial play in US

Melissa Palmer
November 22, 2025

OpenAI is locking in Foxconn as a US-based manufacturing and integration partner for AI data center hardware, covering power, networking,...