Nvidia’s GPUs, specifically the GB300 Blackwell Ultra systems, are now powering Anthropic’s Claude models in Microsoft Azure, following a strategic alliance to expand enterprise access.

For secure deployments, the NVIDIA Secure Agent Workspace Reference Design provides architecture and controls for identity, network access, and runtime policies to protect corporate data as autonomous agents operate across networks.

Enterprises can run Claude via their existing Azure accounts, benefiting from in-situ authentication, billing, networking, governance, and data controls, with inference processed in Azure and data residency options available.

GuruFocus gives Nvidia a GF Score of 95/100, praising profitability and growth while cautioning on valuation and insider selling concerns.

Valuation indicators show a P/E around 29.7x, reflecting expectations of continued earnings growth driven by AI expansion.

The production barrier isn’t just model quality but the surrounding logistics—procurement, governance, security, and operation within existing cloud environments.

Anthropic’s rapid 2026 growth includes Claude 4.x, the Fable 5 launch, and a high-valuation Series H that pushes its market value toward the hundreds of billions.

Nvidia’s $10 billion investment and GPU infrastructure provision creates a flywheel: more compute drives more Nvidia deployments, fueling revenue growth.

Anthropic aims to consume up to 1 gigawatt of computing power, signaling a scale-at-center operation alongside data-center operators and energy players.

The rollout has faced challenges, including temporary disablement of Fable 5 due to export controls and controversy over alleged illicit use of Claude technology via fake accounts.

The deal contemplates up to $15 billion in total investment into Anthropic: up to $5 billion from Microsoft and up to $10 billion from Nvidia, lifting Anthropic’s valuation to about $350 billion.

Nvidia, valued around $4.7 trillion with its stock near $194, is a central enabler for large language models and enterprise AI workloads.