NVIDIA Reports 92% Data Centre Growth, Unveils VeraRubin Roadmap in Q1 2027

NVIDIA Corporation reported financial results for the first quarter of 2027, revealing a 92 per cent year-on-year increase in data centre revenue. The company attributed this growth to the rapid transition of hyperscale workloads from CPU to GPU-based accelerated computing, a shift driven by an inflection in inference demand. Management noted that Blackwell systems experienced the fastest product ramp in the company’s history.

To better reflect diverse growth drivers, NVIDIA introduced a new reporting framework that splits its market into 'Hyperscale' and 'ACIE' (AI Clouds, Industrial, and Enterprise) segments. The ACIE segment is designed to capture expansion in AI natives, sovereign clouds, and industrial sectors, positioning NVIDIA’s integrated, full-stack solutions as alternatives to custom chip design. Management stated that while hyperscalers build their own stacks, the broader market requires fully integrated platforms they can operate without designing their own chips.

The company provided guidance for the second quarter, expecting sequential growth driven primarily by data centre demand. However, NVIDIA explicitly excluded any potential revenue from China in its outlook due to regulatory uncertainty, despite holding approved licences for H200 chips. Operating expenses are projected to grow in the upper forties percentage range for the full year, driven by intensified research and development and the internal usage of AI productivity tools.

NVIDIA outlined its roadmap for the VeraRubin platform, with production shipments scheduled to commence in the second half of the year, starting in the third quarter. Management expressed confidence in achieving $1 trillion in combined Blackwell and Rubin revenue between 2025 and calendar 2027. Jensen Huang asserted that every major frontier model company is adopting the VeraRubin architecture from the outset, noting the company is gaining share in AI inference as its architecture supports the entire lifecycle from pre-training to agentic orchestration.

In addition to operational updates, NVIDIA announced a quarterly dividend increase from $0.01 to $0.25 per share and authorised an $80 billion share repurchase program. The company also highlighted emerging markets beyond traditional data centres, with physical AI and robotics revenue exceeding $9 billion over the last 12 months. NVIDIA expects to become a leading CPU supplier, with visibility to nearly $20 billion in total CPU revenue this year driven by the new Vera architecture, designed as an 'agentic CPU' for orchestration.

Long-term projections suggest AI infrastructure spending could reach $3 trillion to $4 trillion annually by the end of the decade as Agentic AI proliferates. Management believes the ACIE segment will eventually surpass the hyperscale segment in size due to the scale of the global industrial economy. The company’s strategic positioning centres on an 'extreme codesign' approach, aiming to deliver the industry’s lowest token cost and highest return on investment for AI factory operators.