AI · SEMICONDUCTOR · FACT CHECK
SKD WIRETrue
AI cloud company Nebius has acquired Enterprise, a move aimed at addressing GPU cold-start delays in its cloud infrastructure and reinforcing the company's inference software stack, according to reporting based on primary sources.
Cold-start latency — the wait time required to spin up GPU instances before inference workloads can run — has been a persistent pain point for AI cloud providers serving latency-sensitive applications. The acquisition of Enterprise is intended to eliminate this bottleneck, allowing Nebius customers to bring GPU capacity online more quickly for inference workloads.
Beyond hardware provisioning, the deal is positioned as a reinforcement of Nebius's inference stack, the software layer that serves large language model (LLM) and other AI workloads in production. By integrating Enterprise's capabilities, Nebius aims to offer a more complete offering for enterprises deploying AI models at scale.
The move aligns with broader demand for inference-optimized cloud services as companies shift from model training to deployment. Nebius, which operates GPU cloud infrastructure, is competing with larger hyperscalers and AI-focused cloud providers for enterprise inference workloads.
Verdict: True
