Nscale Buys Anyscale To Bolt Ray-Powered Compute Onto Sovereign AI Cloud

UK AI cloud provider Nscale has agreed to acquire San Francisco distributed-compute startup Anyscale, folding the Ray creators into its European sovereign AI infrastructure push.

Nscale Buys Anyscale To Bolt Ray-Powered Compute Onto Sovereign AI Cloud

Nscale, the London-based AI cloud provider that has spent the past year positioning as Europe's sovereign alternative to US hyperscalers, has agreed to acquire San Francisco-based distributed-compute startup Anyscale. The deal, disclosed on July 30, 2026, folds the creators of the open-source Ray framework into Nscale's Nordic and UK GPU cluster footprint.

What Anyscale Brings

Anyscale was founded in 2019 by the UC Berkeley researchers behind Ray, the distributed compute framework that has become a de facto standard for scaling Python-based AI workloads across large GPU clusters. Its platform helps enterprises run data processing, training, fine-tuning, reinforcement learning and agentic-inference pipelines on any cloud. Anyscale had raised a $99M Series C in December 2021 at a $1B+ valuation, followed by strategic partnerships with Microsoft Azure (June 2026) and NVIDIA.

What Nscale Gains

Nscale's business is capacity: purpose-built AI data centres in the UK, Norway and elsewhere, backed by long-term power procurement. But raw GPUs are not what enterprise buyers actually consume — they consume a scheduler, a training runtime and an inference layer. Anyscale plugs those pieces directly into Nscale's stack, turning Nscale from an infrastructure host into a full-stack sovereign AI cloud in one transaction.

Nscale AI data center campus

The Sovereign AI Backdrop

Nscale is one of the loudest voices in Europe's push to build sovereign AI capacity independent of US hyperscalers. Earlier this summer the company closed a $900M revolving credit facility to accelerate global data center construction, on top of the $14.6B valuation it hit in June via a large equity round. Owning the compute framework used by the world's largest AI trainers strengthens the pitch.

Deal Terms And Timing

Terms of the acquisition were not disclosed. Both parties said the transaction is subject to customary regulatory approvals. The move joins a busy stretch of AI-infrastructure consolidation — see our recent coverage of the Samsung-Broadcom $200B memory and foundry pact, Microsoft Azure crossing $100B in annual revenue and AWS's $200B Q2 print.

Reporting based on coverage from Tech Startups and industry filings.

Category: M&A

Tags: venture capital funding Enterprise AI artificial intelligence Mergers & Acquisitions AI Infrastructure

Related Articles