1. Reporting Framework Overhaul and the ACIE Revelation. NVIDIA replaced its prior segment structure with two platforms: Data Center (Hyperscale and ACIE) and Edge Computing. The analytically significant finding is that ACIE — AI-native clouds, industrial deployments, and sovereign AI — generated $37B in Q1 FY2027, nearly matching Hyperscale's $38B. ACIE grew 31% sequentially versus 12% for Hyperscale. Sovereign AI revenue grew over 80% YoY and is now deployed in approximately 40 countries representing $50T in GDP. The number of partner data centers exceeding 10MW nearly doubled in one year to over 80 sites. Management posted nine quarters of historical revenue under the new framework on its investor relations website.¹

2. Vera CPU Entry into a New $200B TAM. Management disclosed the Vera CPU — a custom Arm-based processor co-designed end-to-end with Rubin GPUs and NVLink — and provided specific FY2027 revenue visibility of nearly $20B in CPU revenue. Jensen Huang's prepared remarks described this as "a $200 billion TAM we have never addressed before." Every major hyperscaler and system maker was cited as a launch partner. VeraRubin production shipments begin in Q3 FY2027, integrating seven purpose-built chips across five rack types to deliver up to 35x higher inference throughput than Blackwell. The $20B CPU revenue figure, if achieved, would add a product category equivalent to approximately one quarter of Edge Computing's current annual run rate from a standing start.¹

3. Capital Return Step-Change. NVIDIA returned a record $20B to shareholders through buybacks in Q1. The quarterly dividend was raised 25-fold from $0.01 to $0.25 per share — a figure Jensen Huang corrected live on the call after CFO Colette Kress initially stated $0.20. A new $80B repurchase authorization was added on top of $39B remaining on the existing plan. Management committed to returning approximately 50% of free cash flow for the full year. At the Q1 FCF run rate of $48.6B, 50% implies roughly $24B per quarter in shareholder returns going forward.¹²

4. Anthropic as a New Strategic Compute Partner. NVIDIA disclosed a strategic partnership with Anthropic to expand its compute capacity across AWS, Azure, CoreWeave, and other providers. Management explicitly noted prior NVIDIA coverage of Anthropic compute was "largely zero." Anthropic joins OpenAI — whose GPT 5.5 was co-designed with Blackwell and launched this quarter — xAI, Meta, Google DeepMind, Perplexity, and Cursor as NVIDIA-infrastructure-dependent frontier labs. The partnership adds a previously uncovered frontier AI demand source to NVIDIA's customer base.¹

NVIDIA Q1 FY2027: $81.6B, Record Free Cash Flow, and a New $200B CPU Bet

The Power Delivery Gap