Does NVIDIA aim to grow faster than hyperscaler CapEx, and do you expect hyperscaler CapEx to continue rapid growth beyond this year?

Jensen affirmed that NVIDIA should grow faster than hyperscaler CapEx due to its diverse Data Center segments, particularly the rapidly expanding ACIE segment (AI natives, enterprise, industrial, sovereign AI). He expects hyperscaler CapEx to continue growing significantly, driven by the fundamental shift to AI-centric computing, where compute directly translates to revenue and profit. “First of all, we should be growing faster than hyperscale CapEx. And the reason for that is illustrated by the segmentation that I just described.”

How will Vera Rubin and extreme co-engineering impact NVIDIA's share of the inference market in late '26 and '27?

Jensen stated NVIDIA is rapidly gaining inference share, especially with new frontier model companies and partnerships like Anthropic. He expects Vera Rubin to be even more successful than Grace Blackwell, with every frontier model company adopting it from the outset. He highlighted NVIDIA's unique position in serving the fragmented ACIE segment and physical AI, where its inference share is already dominant. “Vera Rubin is going to be even more successful than Grace Blackwell at this point. Every single, I can't think of one, every single frontier model company will jump on Vera Rubin from the get-go, and that wasn't true before on Blackwell.”

What traction is NVIDIA seeing with custom merchant chips like LPX, and how does it fit into the broader platform strategy?

Jensen described LPX as a niche product designed for low latency and high token rate, but with limited throughput and model size capacity due to its SRAM-based architecture. He reiterated that Grace Blackwell and Vera Rubin are the primary platforms for the entire AI lifecycle (data processing, training, inference), while LPX serves specific premium token services where high interactivity is paramount. “LPX and other SRAM-based decode focused -- high token rate generated focused accelerators will always be -- would be a niche product for some time to come.”

NVDA

NVIDIA (NVDA) Q1 FY27 earnings call: guidance and Q&A

Earnings call· Apr 2026 (Q1 FY27)· Information Technology· May 20, 2026

Source: earnings call transcript

Executive summary

NVIDIA Q1 FY27 — Blackwell Ramp and Vera CPU Launch Drive Record Results

NVIDIA delivered an extraordinary Q1 FY27, driven by the rapid adoption of its Blackwell architecture and the emergence of agentic AI. The company achieved record revenue and free cash flow, while also significantly increasing its capital return to shareholders. Management highlighted the successful ramp of Blackwell systems across diverse customer bases and the strategic entry into the CPU market with Vera, positioning NVIDIA at the center of the evolving AI and physical AI landscape.

Highlights

Total revenue reached a record $82 billion, up 85% year-over-year and 20% sequentially.
Data Center revenue was $75 billion, increasing 92% year-over-year and 21% sequentially, driven by Blackwell architecture.
Generated record free cash flow of $49 billion, up from $35 billion in Q4 FY26.
Increased quarterly dividend from $0.01 to $0.25 per share, demonstrating confidence in long-term free cash flow.
Announced a new $80 billion share repurchase authorization, adding to the $39 billion remaining on the current plan.

Concerns

Uncertainty regarding H200 revenue from China-based customers, with no revenue included in the outlook due to import restrictions.
Consumer demand in Edge Computing fell modestly due to higher memory and system prices.
Days sales outstanding (DSO) is expected to return to the mid-50s in Q2 FY27, up from 45 days in Q1 FY27.

Guidance & targets

Category	Target	Timeline	Confidence
Total Revenue	$91 billion, plus or minus 2% high materiality	Q2 · FY27	High
GAAP Gross Margin	74.9%, plus or minus 50 basis points medium materiality	Q2 · FY27	High
Non-GAAP Gross Margin	75%, plus or minus 50 basis points medium materiality	Q2 · FY27	High
Full-year Gross Margins	Mid-70s medium materiality	FY27	High
GAAP Operating Expenses	Approximately $8.5 billion medium materiality	Q2 · FY27	High
Non-GAAP Operating Expenses	Approximately $8.3 billion medium materiality	Q2 · FY27	High
Full-year OpEx Growth	Upper 40s on a year-over-year basis medium materiality	FY27	High
Full-year GAAP and Non-GAAP Tax Rates	Between 16% and 18% low materiality	FY27	High
Blackwell and Rubin Revenue	$1 trillion high materiality	2025 through calendar 2027	High
Share Repurchase Authorization	$80 billion high materiality	Ongoing	High
Capital Return to Shareholders	Roughly 50% of free cash flow high materiality	This year · FY27	High
Quarterly Dividend	$0.25 per share high materiality	Ongoing	High
Vera Rubin Production Shipments	Commence in the second half of this year, starting in Q3 high materiality	Q3	High
Vera CPU Revenue	Nearly $20 billion high materiality	This year · FY27	High
AI Infrastructure Spending	$3 trillion to $4 trillion annually high materiality	By the end of this decade	High

Segment performance

Segment	Revenue	YoY	QoQ	Margin
Data Center Driven by sustained strength in Blackwell architecture and demand for GB300 NVL72. InfiniBand grew more than 4x year-over-year. Computing Revenue: $60 billionComputing Revenue Growth YoY: 77%Networking Revenue: $15 billionNetworking Revenue Growth YoY: nearly tripled	$75 billion	92%	21%	—
Data Center - Hyperscale Includes revenue from public cloud and the world's largest consumer Internet companies. Percentage of Data Center Revenue: approximately 50%	$38 billion	—	12%	—
Data Center - ACIE Incorporates AI clouds, industrial, and enterprise. NVIDIA AI infrastructure is now deployed across nearly 40 countries. AI Cloud Revenue Growth YoY: more than tripledSovereign Revenue Growth YoY: 80%	$37 billion	—	31%	—
Edge Computing Robust Blackwell workstation demand was a strong contributor. Consumer demand fell modestly due to higher memory and system prices.	$6.4 billion	29%	10%	—

Product announcements

Product	Type	Timeline	Details
Vera Rubin	launch	Second half of this year, starting in Q3
Vera CPU	launch	This year

Deals & partnerships

OpenAI Collaboration

GPT-5.5 was codesigned for, trained with, and served on Blackwell. OpenAI's Codex is a harness around GPT-5.5 model.

Microsoft Collaboration

Microsoft's Fairwater, the world's most powerful AI data center, is now live, ahead of schedule, powered by hundreds of thousands of Blackwell GPUs.

AWS Collaboration

Starting this year, AWS will add more than 1 million Blackwell and Rubin GPUs and are collaborating on Spectrum Networking.

Google Collaboration

Blackwell will be offered to customers in the cloud, including confidential computing capability. Google's A5X bare metal instances can support up to 960,000 Rubin GPUs across multiple sites.

Anthropic Strategic Partnership

Deepened collaboration and strategic partnership to expand their compute capacity through AWS, Azure, CoreWeave, SpaceXAI and more.

Uber Partnership by 2028

Partnership will power the robotaxi fleet across nearly 30 cities and 4 continents by 2028.

Cadence, Synopsys, Siemens, Adobe Collaboration

Working closely to accelerate all of the world's tools to run on CUDA, as agents use these tools.

Risks & headwinds

China H200 Revenue Uncertainty Q2 FY27 and beyond

No revenue included in outlook

Mitigation:Management is uncertain if any imports will be allowed into the country, and has not included any China data center compute revenue in its outlook.

Consumer Demand Decline in Edge Computing Q1 FY27

Modest fall

Mitigation:Attributed to higher memory and system prices.

Supply Challenges Ongoing

Not immune

Mitigation:Company is confident in its ability to support growth opportunities through intense focus, scale, and long-standing partnerships with critical suppliers, increasing total supply to $145 billion.

Q&A highlights

What is the philosophy behind the new segmentation, competitive differences between segments, and the role of the new CPU in these segments?

Jensen explained the new segmentation (Hyperscale, ACIE, Edge Computing) reflects the diverse nature of AI applications, deployment environments, and go-to-market strategies. He emphasized NVIDIA's unique full-stack, co-designed approach that caters to these varied needs, from hyperscalers to enterprises and robotic systems. The CPU plays a role in agentic AI orchestration.

“NVIDIA is quite unique in the sense that we are the only company that builds all of the technology components. We build it in an extreme co-design way in a complete end-to-end way, in a full stack way.”

asked by Joseph Moore · answered by Jen-Hsun Huang

2 min read 6 chapters

Detailed narrative

New Reporting Framework for Market Platforms

NVIDIA introduced a new reporting framework to better reflect its current and future growth drivers, segmenting its business into two market platforms: Data Center and Edge Computing. Within Data Center, two submarkets are now reported: Hyperscale (public cloud and large consumer internet companies) and ACIE (AI clouds, industrial, and enterprise).

This new structure aims to provide clearer insights into the diverse applications and deployment environments for NVIDIA's technology.

Blackwell Architecture's Rapid Adoption

The Blackwell architecture experienced the fastest product ramp in NVIDIA's history, with GB300 NVL72 seeing strong demand from frontier model builders and hyperscalers. Key deployments include Microsoft's Fairwater AI data center, powered by hundreds of thousands of Blackwell GPUs, and AWS's plan to add over 1 million Blackwell and Rubin GPUs. This widespread adoption underscores Blackwell's position as a leading AI computing platform.

Strategic Entry into the CPU Market with Vera

NVIDIA is making a significant entry into the CPU market with Vera, a CPU purpose-built for agentic AI. This initiative opens a new $200 billion Total Addressable Market (TAM) for the company, a segment it had not previously addressed. Vera is designed to deliver up to 1.5x faster performance per core, 2x performance per watt, and 4x density per rack compared to x86-based alternatives, with production shipments slated to begin in Q3.

Expanding AI Infrastructure Market Opportunity

The build-out of AI factories is accelerating at an unprecedented pace, with AI infrastructure spending projected to reach $3 trillion to $4 trillion annually by the end of the decade. This growth is fueled by the transition of hyperscale workloads to GPU-based computing and the inflection in adoption of AI-native products and services.

The value of NVIDIA AI infrastructure is rising, as evidenced by H100 cloud pricing increasing 20% year-to-date.

Agentic AI and Physical AI Momentum

Agentic AI, characterized by reasoning and autonomous actions, is driving new growth opportunities, with Vera CPU designed to handle orchestration while GPUs perform the core 'thinking.' Beyond data centers, NVIDIA is seeing significant momentum in physical AI, which generated over $9 billion in revenue in the last 12 months.

Partnerships, such as with Uber for robotaxi fleets, highlight the expansion into autonomous and robotic systems operating in the physical world.

Supply Chain Management and Capital Allocation Strategy

NVIDIA increased its total supply, including inventory purchase commitments and prepaid, to $145 billion to meet the incredible demand. The company's capital allocation strategy prioritizes R&D and strategic investments to cultivate its ecosystem and strengthen market position. This is complemented by a substantial increase in its quarterly dividend to $0.25 per share and a new $80 billion share repurchase authorization, reflecting confidence in long-term free cash flow.

Summary of the company's earnings call; see how call summaries are prepared. Not investment advice.