AI Compute

Twitter

Facebook

Copy Link

Key Facts

Key PlayersNVIDIA, Microsoft, Google, AWS, CoreWeave

Top News

Nvidia weighs $10 billion investment in Anthropic IPO as AI firm targets historic $2 trillion valuation

Nvidia is in talks to invest up to $10 billion in Anthropic's planned IPO, which seeks to raise $100 billion at a $2 trillion valuation. The chipmaker's potential anchor investment highlights deepening ties between AI hardware suppliers and model developers, as Anthropic's revenue surged to $65 billion annualized by July 2025.

Business6 sources

Nvidia weighs $10 billion investment in Anthropic IPO as AI firm targets historic $2 trillion valuation

Business6 sources

Oracle Deploys 850 MW AI Capacity, Unveils Agent-Driven SaaS Strategy to Counter Industry Disruption

Oracle co-CEOs Mike Sicilia and Clay Magouyrk outlined an aggressive AI expansion during Q1 FY 2027 earnings, delivering 850 MW of datacenter capacity and over 300,000 GPUs. The company closed $30 billion in new AI contracts while introducing an agentic AI accelerator designed to compress SaaS deployments from years to weeks, positioning AI agents as workflow enhancers rather than application replacements.

Technology2 sources

Oracle Deploys 850 MW AI Capacity, Unveils Agent-Driven SaaS Strategy to Counter Industry Disruption

Technology2 sources

Z.AI Launches $5 Billion Fundraising Through Hong Kong Share Sale and Convertible Bonds

Beijing-based AI developer Z.AI has launched a massive $5 billion capital raise through a dual offering in Hong Kong—a $2 billion share placement and $3 billion convertible bond sale. The company plans to use proceeds for AI research and development, computing resources, infrastructure expansion, and strategic acquisitions.

Business2 sources

Z.AI Launches $5 Billion Fundraising Through Hong Kong Share Sale and Convertible Bonds

Business2 sources

UAE Revises $30bn AI Data Center Plan After Iranian Attacks Target Critical Infrastructure

The UAE is overhauling its ambitious 5-gigawatt AI data center project after Iranian attacks damaged regional infrastructure in March. Originally planned as a single 10-square-mile campus in Abu Dhabi, the project will now be distributed across multiple sites with underground facilities and enhanced air defenses to protect against drone and missile threats.

Business5 sources

UAE Revises $30bn AI Data Center Plan After Iranian Attacks Target Critical Infrastructure

Business5 sources

OpenAI Halts $200 Pro Subscriptions as Unprecedented Astra Demand Strains Infrastructure

OpenAI pauses new Pro subscriptions after overwhelming demand for its newest AI model Astra pushes systems to their limits. The company's $200-per-month tier remains unavailable as OpenAI works to add compute capacity while maintaining service quality for existing users.

Technology5 sources

OpenAI Halts $200 Pro Subscriptions as Unprecedented Astra Demand Strains Infrastructure

Technology5 sources

d-Matrix Joins Nvidia's AI Infrastructure Ecosystem with NVLink Fusion Integration for Raptor XPU

AI inference specialist d-Matrix announced integration of Nvidia's NVLink Fusion chip-linking technology and MGX rack designs into its upcoming Raptor XPU platform. The collaboration enables d-Matrix to deploy its specialized inference chips within Nvidia's AI factory infrastructure, offering customers a faster path to ultra-low latency AI services while Nvidia expands its walled garden ecosystem beyond GPUs.

Technology7 sources

d-Matrix Joins Nvidia's AI Infrastructure Ecosystem with NVLink Fusion Integration for Raptor XPU

Technology7 sources

HD Hyundai Heavy bets $747 million on power generation engines and SMRs to fuel AI electricity demand

HD Hyundai Heavy Industries announced a 1.07 trillion won ($747 million) investment to build new production facilities for power-generation engines and small modular reactors in Ulsan. The move targets the rapidly growing AI infrastructure market, with deals already secured from Aperion Energy Group and Corban Energy Group.

Business2 sources

HD Hyundai Heavy bets $747 million on power generation engines and SMRs to fuel AI electricity demand

Business2 sources

AI Data Centers May Create Only 10,400 Jobs by 2035, 75% Below Industry Projections

Economic think tank Verdant challenges the tech industry's rosy job creation forecasts for AI data centers. New analysis suggests UK facilities will generate only 10,400 permanent operational jobs by 2035—just a quarter of the 40,200 roles projected by industry body techUK. The discrepancy stems from automation in newer, larger facilities requiring far fewer staff than older sites.

Business2 sources

AI Data Centers May Create Only 10,400 Jobs by 2035, 75% Below Industry Projections

Business2 sources

NVIDIA CUDA Toolkit 13.4 Brings Windows on Arm Support, Preps AI Developers for RTX Spark Launch

NVIDIA released CUDA Toolkit 13.4, marking the first time CUDA officially extends native Arm development from Linux to Windows. AI developers can now compile CUDA applications directly on Windows Arm64 systems and gain early developer access to the Vera Rubin architecture ahead of RTX Spark's October launch.

Technology2 sources

NVIDIA CUDA Toolkit 13.4 Brings Windows on Arm Support, Preps AI Developers for RTX Spark Launch

Technology2 sources

Nvidia Partners with 8 Australian Data Centres to Build 2GW AI Infrastructure by 2027

Nvidia announced partnerships with eight Australian data centre operators to build AI factory capacity targeting 2 gigawatts by 2027. The expansion could more than double Australia's current 1.6GW computing capacity but raises questions about power supply and sovereignty as the chipmaker's energy demands will exceed all existing data centres combined.

Technology6 sources

Nvidia Partners with 8 Australian Data Centres to Build 2GW AI Infrastructure by 2027

Technology6 sources

Dell CEO Michael Dell warns structural AI compute shortage will worsen in 2027

Dell Technologies CEO Michael Dell warned that the structural shortage in AI-driven compute will intensify in 2027, surpassing 2026 constraints. Speaking at Goldman Sachs Communacopia + Technology Conference, Dell revealed the company holds $9.5 billion in AI server backlog while semiconductor fabrication plants struggle to meet demand from accelerating AI adoption across enterprises.

Business2 sources

Dell CEO Michael Dell warns structural AI compute shortage will worsen in 2027

Business2 sources

Apple's A20 Pro Chip Powers iPhone 18 Pro with 2nm Tech and Enhanced On-Device AI

Apple unveiled the A20 Pro chip, the world's first 2nm smartphone processor, designed to power the iPhone 18 Pro lineup. The chip delivers 40% faster graphics, 20% CPU performance gains, and doubled Neural Engine compute with advanced thermal management for sustained AI performance.

Technology10 sources

Apple's A20 Pro Chip Powers iPhone 18 Pro with 2nm Tech and Enhanced On-Device AI

Technology10 sources

AI Token Cost Plummets 41% as Enterprises Shift Away from Premium Models

The price businesses pay for AI tokens has dropped 41% since March 2024, falling from $1.15 to 68 cents per million tokens. Companies are moving away from expensive frontier AI models as cost discipline replaces the earlier "tokenmaxxing" trend, threatening the growth assumptions behind OpenAI and Anthropic's blockbuster valuations.

Business3 sources

AI Token Cost Plummets 41% as Enterprises Shift Away from Premium Models

Business3 sources

Nebius Auctions AI Compute Capacity at 20% Premium as Demand Outpaces Supply

Nebius Group turned to auctions for scarce Nvidia Blackwell AI compute capacity, achieving prices 15-20% above previous highs as demand continues to outstrip supply. The AI infrastructure provider is deploying customer prepayments and asset-backed financing to fund its $20-25 billion capital expenditure program while expanding contracted power capacity to 5 gigawatts.

Business3 sources

Nebius Auctions AI Compute Capacity at 20% Premium as Demand Outpaces Supply

Business3 sources

HP Inc partners with NVIDIA and Red Hat to launch Enterprise AI Platform with 20 PFLOPS performance

HP Inc announced a collaboration with Red Hat and NVIDIA to develop an enterprise-grade AI platform for production inference. The solution combines HP ZGX Fury powered by NVIDIA GB300 Grace Blackwell Ultra Desktop Superchip with Red Hat AI Factory, delivering up to 20 PFLOPS FP4 AI performance for Local AI inference while addressing latency, Privacy, and Data sovereignty requirements.

Technology2 sources

HP Inc partners with NVIDIA and Red Hat to launch Enterprise AI Platform with 20 PFLOPS performance

Technology2 sources

Prev12 3 4…89 Next