NVDA · Q1 2027 Earnings
BullishNvidia
Reported May 20, 2026
30-second summary
Revenue of $81.6B (+85% YoY, +20% QoQ) cleared the $78.0B guide midpoint by $3.6B and the high end by $2.1B, with Data Center at $75.2B (+92% YoY) and non-GAAP EPS of $1.87. Q2 FY2027 is guided to $91.0B (±2%), implying ~95% YoY growth at midpoint (range +91–99%) against the $46.74B prior-year base — a step that puts to bed any "Blackwell plateau" thesis and sets up Vera Rubin production shipments to begin Q3. Management restructured Data Center reporting into Hyperscale vs. ACIE (AI Clouds, Industrial, Enterprise), with ACIE up 31% QoQ — a deliberate move to redirect the growth narrative away from hyperscaler concentration. Capital returns stepped up sharply: $20B returned in Q1, a new $80B repurchase authorization, and the quarterly dividend raised from $0.01 to $0.25 per share.
Headline numbers
EPS
Q1 FY2027
$1.87
Revenue
Q1 FY2027
$81.61B
+85.0% YoY
Gross margin
Q1 FY2027
74.9%
Free cash flow
Q1 FY2027
$48.55B
Operating margin
Q1 FY2027
65.6%
Key financials
Q1 FY2027| Metric | Q1 FY2027 | YoY | Q4 FY2026 | QoQ |
|---|---|---|---|---|
| Revenue | $81.61B | +85.0% | $68.13B | +19.8% |
| EPS | $1.87 | — | $1.62 | +15.4% |
| Gross margin | 74.9% | — | 75.2% | -30bps |
| Operating margin | 65.6% | — | 67.7% | -210bps |
| Free cash flow | $48.55B | — | $34.90B | +39.1% |
Guidance
Guidance is issued one quarter forward. The Prior-guide column references the guide issued last quarter for the period just reported; the New-guide column is for next quarter.
Actuals vs prior guidance
| Metric | Period | Prior guide | Actual | Δ | Result |
|---|---|---|---|---|---|
| Revenue | Q1 FY2027 | $78.0B ± 2% | $81.615B | +$3.6B above guidance (midpoint) | Beat |
| Gross Margin (GAAP) | Q1 FY2027 | 74.9% ± 50 bps | 74.9% | in-line | Met |
| Gross Margin (Non-GAAP) | Q1 FY2027 | 75.0% ± 50 bps | 75.0% | in-line | Met |
| Operating Expenses (GAAP) | Q1 FY2027 | approximately $7.7B | Not disclosed | — | Beat |
| Operating Expenses (Non-GAAP) | Q1 FY2027 | approximately $7.5B | Not disclosed | — | Beat |
New guidance
| Metric | Period | Guide | YoY |
|---|---|---|---|
| Revenue | Q2 FY2027 | $91.0B ± 2% | +95-97% YoY |
| Gross Margin (GAAP) | Q2 FY2027 | 74.9% ± 50 bps | — |
| Gross Margin (Non-GAAP) | Q2 FY2027 | 75.0% ± 50 bps | — |
| Operating Expenses (GAAP) | Q2 FY2027 | $8.5B | — |
| Operating Expenses (Non-GAAP) | Q2 FY2027 | $8.3B | — |
Changes to prior guidance
| Metric | Period | Prior guide | New guide | Δ | Result |
|---|---|---|---|---|---|
| Tax Rate (GAAP and Non-GAAP) | FY 2027 | 17.0% to 19.0% | 16.0% to 18.0% | -100 bps at both low and high end | Lowered |
Reaffirmed unchanged this quarter: Gross Margin (mid-70s)
Segment performance
Q1 FY2027| Segment | Q1 FY2027 | YoY |
|---|---|---|
| Data Center | $75.2B | +92.0% |
| Edge Computing | $6.4B | +29.0% |
| Data Center - Compute | $60.4B | +77.0% |
| Data Center - Networking | $14.8B | +199.0% |
| Data Center YoY Growth | 92% | — |
| Data Center Networking YoY Growth | 199% | — |
| Data Center Compute YoY Growth | 77% | — |
| Edge Computing YoY Growth | 29% | — |
Profitability
Q1 FY2027| Segment | Q1 FY2027 |
|---|---|
| Gross Margin (Non-GAAP) | 75.0% |
| Operating Margin (GAAP) | 65.6% |
| Free Cash Flow | $48.6B |
| FCF Margin | 59.5% |
Management tone
Q2 FY2026 Blackwell ramp → Q3 FY2026 $500B is a floor → Q4 FY2026 compute equals revenues → Q1 FY2027 AI factory economics and ACIE diversification
The defining shift this quarter is structural: management restructured Data Center reporting to bifurcate Hyperscale from ACIE (AI Clouds, Industrial, Enterprise). "We have two market platforms, Data Center and Edge Computing. Within Data Center, we will report to sub-markets, Hyperscale and ACIE… ACIE revenue was 37 billion and grew 31% quarter over quarter, including AI cloud revenue that more than tripled year over year." Three quarters ago Jensen was arguing that demand was durable; two quarters ago he framed compute as the revenue function of every customer; this quarter the reporting structure itself is engineered to direct investor attention toward the faster-growing, more diversified second platform. The reshuffle signals confidence that the next leg of growth is not hyperscaler-concentrated.
The supply posture has fully inverted from the Q2 FY2026 "supply constraints persist" framing. Last quarter management cited "strategically secured inventory and capacity"; this quarter they quantified it: "We increased total supply, inclusive of inventory, purchase commitments on prepaids to $145 billion… giving us full confidence in the 1 trillion in Blackwell and Rubin revenue we foresee from 2025 through calendar 2027." Supply has moved from binding constraint to validated capacity backing a stated TAM number — and the $1T figure is now framed as a base case rather than a stretch.
Vera Rubin was reframed from "next-generation product" to "next-generation stack integration with a new CPU TAM." Vera CPU is now a standalone $200B TAM in management's framing — a brand-new market they have never addressed. "Vera CPU opens a brand new 200 billion TAM for NVIDIA, a market we have never addressed before. We have visibility to nearly 20 billion in total CPU revenue this year." The willingness to disclose a discrete near-term CPU revenue line ($20B standalone) and frame it as additive to Vera Rubin pairing (not cannibalistic) is the most expansionist platform claim Jensen has made in the cycle.
The customer-value framing has hardened into a fully articulated TCO argument: "Customers do not buy GPUs. They build AI factories. And the right economic metric is not the purchase price of the GPU. It is the lifetime cost of an AI factory producing intelligence, token per watt, tokens per dollar, uptime, utilization, time to production, software durability, and asset life." Two quarters ago the moat was framed around performance-per-watt; this quarter it's reframed around total lifetime factory economics — a more durable defense against ASIC commoditization narratives and a direct setup for Vera Rubin's claimed "35x higher inference throughput and 10x greater AI factory revenue compared with Blackwell."
China continues to be excluded from outlook with explicit uncertainty: "While the U.S. government has approved licenses for H-200 to be shipped to China-based customers, we have yet to generate any revenue, and we are uncertain whether any imports will be allowed into the country." The posture is unchanged from Q4 — license approved, revenue zero, outlook excludes — meaning China remains a free option.
Recurring themes management leaned on this quarter:
Risks management surfaced:
Q&A highlights
Joseph Moore · Morgan Stanley
What drove the change in segmentation philosophy? Can you explain competitive differences between segments and discuss the surprising CPU numbers across both segments?
Management explained the new three-segment structure (hyperscale clouds, AI natives/enterprise/sovereign AI, and robotic edge) is designed to help investors understand business diversity across industries, applications, deployment locations, and governance models. Emphasized NVIDIA's unique ability to provide full-stack integrated solutions. AI is diverse across manufacturing, life sciences, energy, and other sectors; applications span enterprise to edge; deployment ranges from hyperscale to on-premise to industrial.
Ben Reitzes · Melius Research
Should NVIDIA grow faster than hyperscaler CapEx? Are you comfortable endorsing that view? Will hyperscaler CapEx continue growing rapidly after this year?
Management reaffirmed expectation to grow faster than hyperscaler CapEx growth (forecasted at 90-100%). Explained that hyperscale CapEx will continue growing toward $3-4 trillion by decade end because compute drives revenue and profit in AI era. However, emphasized the second category (AI natives, enterprises, industrial, sovereign AI) is growing even faster and represents hundreds of thousands of future companies, providing superior long-term growth opportunity.
Vivek Arya · Bank of America Securities
Is the CPU growth for agentic applications incremental or cannibalistic to GPU workloads? Is the $20 billion standalone CPU number included in Vera Rubin or separate?
Management clarified $20 billion is for standalone Vera CPU, separate from Vera Rubin. Explained Vera has four use cases: standalone CPU, paired with Vera Rubin, with CX9 storage stack, and with CX9 for security/confidential computing. Described agentic workloads as complementary: CPU handles orchestration/tool use (browsers, compilers, simulators) while GPUs handle AI inference/thinking. Anticipates billions of future agents each using CPU-based tool interfaces, making CPU growth incremental and additive. Expects to be supply-constrained on Vera throughout its lifecycle.
Timothy Arcuri · UBS
Can you discuss traction with custom merchant chips (CPX, LPX) and how LPX fits into broader platform strategy given historical 20% fast inference market?
Management explained LPX is designed for niche use case of low-latency, high-token-rate inference with limited throughput, model size capacity, and context processing ability. Emphasized LPX use case is intentionally narrow—only for providers with large premium token service portfolios. Expects LPX and SRAM-based accelerators to remain niche products indefinitely. Grace Blackwell and Vera Rubin are positioned as comprehensive platforms supporting full AI lifecycle (data processing, training, post-training, inference). LPX acts as complement only when providers already have high-token-rate service offerings. Current market for such services estimated at less than 20%.
Stacy Raskin · Bernstein Research
Where do AI native clouds fit in the new segmentation? Are you suggesting the AI cloud segment will grow faster than hyperscale, or do you expect similar growth from both?
Management categorized AI native clouds in second segment (alongside enterprise, industrial, sovereign AI) rather than hyperscale. Explained AI natives cannot assemble unrelated parts and have extremely low tolerance for time-to-first-token; NVIDIA architecture is optimal for them. Stated both segments will grow incredibly fast near-term, but long-term expects second category (representing $50-80 trillion global economy) to become larger. Over next few years, expects second category to still grow faster than hyperscale, though both will expand rapidly.
Answers to last quarter's watch list
What to watch into next quarter
Whether Q2 FY2027 revenue clears $92.8B (high end of guide). Q1 beat the guide high by $2.1B; a similar beat would put Q2 at ~$95B and validate the upper end of the +91–99% YoY framing. A guide-midpoint print at $91B would still represent a clean beat against the prior-year base but would be the first quarter without a high-end clear in five.
ACIE quarterly growth rate. ACIE grew +31% QoQ to $37B this quarter, with AI cloud revenue more than tripling YoY. Watch whether ACIE sustains a >20% QoQ pace — that's the engine of the new segmentation narrative, and a deceleration below ~15% QoQ would undercut Jensen's claim that the second platform grows faster than Hyperscale.
Non-GAAP gross margin holding at or above 74.5% as Vera Rubin production begins in Q3. Q2 guide is 75.0% midpoint. The Rubin ramp historically pressures margins; any Q2 print below 74.5% would be the first concrete sign Rubin transition costs are biting earlier than expected.
Vera CPU revenue disclosure cadence. Management gave a near-term CPU visibility figure of "nearly $20B this year" framed as standalone (separate from Vera Rubin). Watch whether Q2 provides a discrete CPU revenue line or remains a Q&A-only metric — disclosed CPU revenue would validate the $200B TAM claim with hard numbers.
Q2 non-GAAP OpEx landing at or near $8.3B. The FY2027 OpEx growth raise to "upper 40s" implies a higher run rate through year-end. Watch whether Q2 actual OpEx lands within $0.2B of guide and whether GAAP operating margin holds above 64%.
China H-200 revenue inclusion. Licenses are approved; revenue is zero; outlook excludes. Any actual H-200 China revenue in Q2 reporting becomes pure upside vs the $91B guide. Watch for any change in management language on import allowance.
Total supply / purchase-commitment trajectory. Total supply reached $145B this quarter, backing the $1T Blackwell+Rubin CY25–CY27 forecast. A Q2 print sustaining $145B+ would reinforce the floor; a step-down would be the first concrete sign forward demand visibility is softening.
Sources
- NVIDIA Q1 FY2027 Press Release, filed with SEC 2026-05-20: https://www.sec.gov/Archives/edgar/data/1045810/000104581026000051/q1fy27pr.htm
- NVIDIA Q1 FY2027 earnings call Q&A (analyst exchanges).
- NVIDIA Q4 FY2026 Press Release and earnings call (for prior-quarter guide baselines and prior-year Q2 FY2026 revenue base).
Get the next brief, free.
We publish analyst-grade earnings briefs the same day or morning after every call — headline numbers, segment KPIs, Q&A highlights, and tone analysis. Free during beta.
This is not investment advice.