Monday, 9 November 2026
Silicon Analytica
Every AI story is a chip story.
All companies
AI accelerators, GPUs and data-center systems

NVIDIA

Latest quarter: Q2 FY2027, ended July 2026

NVDA · NasdaqDesign & fabless

Compare in Financials

About the company

NVIDIA Corporation designs graphics processing units (GPUs), networking products, systems and software for accelerated computing. It reports two segments. Compute & Networking includes data-center GPUs and AI computing platforms, InfiniBand and Ethernet networking, and automotive and robotics platforms. Graphics includes GeForce GPUs for gaming and PCs and professional visualization products. Its CUDA software platform and libraries are central to developer adoption. NVIDIA is fabless: it outsources wafer fabrication, packaging and system assembly to partners such as TSMC. The data center market is now by far its largest source of revenue, driven by demand for AI training and inference.

Silicon Analytica summary based on the company's latest annual report

Key figures

Q2 FY2027 · 26 Jul 2026

Revenue
▲ +105.9% y/y
Operating income
▲ +124.1% y/y
Gross marginCalc.
+2.6 pp y/y
Operating marginCalc.
+5.4 pp y/y
Net income
▲ +125.9% y/y
Free cash flowCalc.
▲ +58.9% y/y

Figures from company filings. Calculated values marked Calc. · Methodology →

How the company earns its revenue

Market-platform view (fiscal 2026 structure). This is a different classification from the reportable segments; do not add them together.

$89.7of every $100 · Data Center$193.7B in FY2026

Total of categories: US$215.9B, reconciles to reported revenue. Different classifications are never added together.

View sources
  • NVIDIA Form 10-K · Notes, revenue by end market (market platform) · FY2026 · published 25 Feb 2026

Quarterly trend

12 available quarters · US GAAP

Revenue, USD (left) · gross margin, % (right) · US GAAPRevenueGross margin$0$50B$100B$150BQ3 FY2024Q1 FY2025Q3 FY2025Q1 FY2026Q3 FY2026Q1 FY2027Revenue, Q3 FY2024: US$18.1B (Reported)$18BRevenue, Q4 FY2024: US$22.1B (Calculated)$22BRevenue, Q1 FY2025: US$26.0B (Reported)$26BRevenue, Q2 FY2025: US$30.0B (Reported)$30BRevenue, Q3 FY2025: US$35.1B (Reported)$35BRevenue, Q4 FY2025: US$39.3B (Calculated)$39BRevenue, Q1 FY2026: US$44.1B (Reported)$44BRevenue, Q2 FY2026: US$46.7B (Reported)$47BRevenue, Q3 FY2026: US$57.0B (Reported)$57BRevenue, Q4 FY2026: US$68.1B (Calculated)$68BRevenue, Q1 FY2027: US$81.6B (Reported)$82BRevenue, Q2 FY2027: US$96.2B (Reported)$96.2B50%90%Gross margin, Q3 FY2024: 74.0% (Calculated)Gross margin, Q4 FY2024: 76.0% (Calculated)Gross margin, Q1 FY2025: 78.4% (Calculated)Gross margin, Q2 FY2025: 75.1% (Calculated)Gross margin, Q3 FY2025: 74.6% (Calculated)Gross margin, Q4 FY2025: 73.0% (Calculated)Gross margin, Q1 FY2026: 60.5% (Calculated)Gross margin, Q2 FY2026: 72.4% (Calculated)Gross margin, Q3 FY2026: 73.4% (Calculated)Gross margin, Q4 FY2026: 75.0% (Calculated)Gross margin, Q1 FY2027: 74.9% (Calculated)Gross margin, Q2 FY2027: 75.0% (Calculated)75.0%Source: NVIDIA quarterly filings and results releases · gross margin = gross profit ÷ revenue, calculated by SiliconAnalyticaSilicon Analyticasiliconanalytica.com/companies/nvidia
View sources
  • NVIDIA Form 10-Q · XBRL us-gaap:Revenues; consolidated statements · Q3 FY2024 · published 20 Nov 2024 · covers
  • NVIDIA Form 10-K · XBRL us-gaap:Revenues; consolidated statements · Q4 FY2024, Q4 FY2025, Q4 FY2026 · published 25 Feb 2026 · covers
  • NVIDIA Form 10-Q · XBRL us-gaap:Revenues; consolidated statements · Q1 FY2025 · published 28 May 2025 · covers
  • NVIDIA Form 10-Q · XBRL us-gaap:Revenues; consolidated statements · Q2 FY2025 · published 27 Aug 2025 · covers
  • NVIDIA Form 10-Q · XBRL us-gaap:Revenues; consolidated statements · Q3 FY2025, Q3 FY2026 · published 19 Nov 2025 · covers
  • NVIDIA Form 10-Q · XBRL us-gaap:Revenues; consolidated statements · Q1 FY2026, Q1 FY2027 · published 20 May 2026 · covers
  • NVIDIA Form 10-Q · XBRL us-gaap:Revenues; consolidated statements · Q2 FY2026, Q2 FY2027 · published 26 Aug 2026 · covers

Financial statements

Full page →

As filed with the SEC (10-K/10-Q), from XBRL data. US$ millions, except per-share amounts; shares in millions.

Line itemQ2 FY2027To 26 Jul 202610-Q ↗Q1 FY2027To 26 Apr 202610-Q ↗Q4 FY2026To 25 Jan 202610-K ↗Q3 FY2026To 26 Oct 202510-Q ↗Q2 FY2026To 27 Jul 202510-Q ↗
Revenue96,22181,61568,127c57,00646,743
Cost of revenue24,07920,45817,034c15,15712,890
Gross profit72,14261,15751,093c41,84933,853
Gross margin75.0%74.9%75.0%73.4%72.4%
Research & development7,0546,3215,512c4,7054,291
Selling, general & administrative1,3541,3001,282c1,1341,122
Operating income63,73453,53644,299c36,01028,440
Operating margin66.2%65.6%65.0%63.2%60.8%
Interest & other, net7,77316,3676,099c1,9262,766
Pre-tax income71,50769,90350,398c37,93631,206
Income tax11,81911,5827,438c6,0264,784
Net income59,68858,32142,960c31,91026,422
Net margin62.0%71.5%63.1%56.0%56.5%
Diluted EPS2.462.39—1.301.08
Diluted shares24,28524,391—24,48324,532

Columns link to the filing. c Calculated by Silicon Analytica from filed figures (hover for the formula); fourth-quarter flows are fiscal year minus nine months. Margins are calculated. Fourth-quarter per-share figures are not filed separately and are left blank.

View sources (5)
  1. NVIDIA 10-Q filed 26 Aug 2026 · accession 0001045810-26-000075
  2. NVIDIA 10-Q filed 20 May 2026 · accession 0001045810-26-000052
  3. NVIDIA 10-K filed 25 Feb 2026 · accession 0001045810-26-000021
  4. NVIDIA 10-Q filed 19 Nov 2025 · accession 0001045810-25-000230
  5. NVIDIA 10-Q filed 27 Aug 2025 · accession 0001045810-25-000209

Values: SEC XBRL company facts (data.sec.gov), refreshed 8 Nov 2026. A period's original filing is used unless a later filing restated it.

Earnings history

NVIDIA's results by period: revenue, margins and segments from stored filings.

Quarters

Fiscal years

Five-year financial history

US GAAP

Latest 5 completed fiscal years, US GAAP, USD, as reported in annual filings (restated comparatives used where later filings revised them). Ratios and growth are Silicon Analytica calculations from the same records.

Revenue

USD · fiscal years · US GAAP · reported

Revenue (USD)FY2022: US$26.9BUS$26.9BFY202230 Jan 2022FY2023: US$27.0BUS$27.0BFY202329 Jan 2023FY2024: US$60.9BUS$60.9BFY202428 Jan 2024FY2025: US$130.5BUS$130.5BFY202526 Jan 2025FY2026: US$215.9BUS$215.9BFY202625 Jan 2026Silicon AnalyticaNVIDIA Form 10-K filed 2026-02-25

Net income attributable to shareholders

USD · fiscal years · US GAAP · reported

Net income attributable to shareholders (USD)FY2022: US$9.8BUS$9.8BFY202230 Jan 2022FY2023: US$4.4BUS$4.4BFY202329 Jan 2023FY2024: US$29.8BUS$29.8BFY202428 Jan 2024FY2025: US$72.9BUS$72.9BFY202526 Jan 2025FY2026: US$120.1BUS$120.1BFY202625 Jan 2026Silicon AnalyticaNVIDIA Form 10-K filed 2026-02-25
NVIDIA annual results
USDFY2022ended 30 Jan 2022FY2023ended 29 Jan 2023FY2024ended 28 Jan 2024FY2025ended 26 Jan 2025FY2026ended 25 Jan 2026
Revenue
Gross profit
Gross margin
Operating income
Operating margin
Net income (attributable)
Diluted EPS
R&D expense
R&D as % of revenue

Business model and industry role

Data-center compute

GPU accelerators (Hopper, Blackwell generations) sold as chips, boards and full rack-scale systems; a GPU is a processor built to run many calculations in parallel, which suits AI workloads.

Networking

InfiniBand and Ethernet switches, network adapters and NVLink interconnects that let thousands of GPUs work as one machine.

Software

CUDA, the programming platform that runs on all NVIDIA GPUs, plus AI libraries, frameworks and enterprise software.

Gaming and professional graphics

GeForce GPUs for PCs and RTX workstation GPUs for design and visualisation.

Automotive

DRIVE computing platforms and software for driver assistance and autonomous driving.

Who buys it

Direct customers are mainly OEMs, ODMs, system integrators and distributors; the end users are cloud providers, AI companies, enterprises and governments 1. NVIDIA reports revenue by the headquarters location of its direct customers, so the geographic split does not show where chips are finally deployed 1.

How it earns revenue

Mostly from hardware sales: chips, boards and complete systems with networking. Since fiscal 2027 NVIDIA groups Data Center revenue into Hyperscale and "AI Clouds, Industrial & Enterprise", and reports Edge Computing separately 3.

What it controls and outsources

NVIDIA controls architecture, chip design, system design and software. Wafer fabrication (TSMC, Samsung), memory (SK hynix, Micron, Samsung), CoWoS advanced packaging and system assembly (Hon Hai, Wistron and others) are supplied by partners 1.

What drives profitabilityIncludes Silicon Analytica view

Gross margin was 71.1% in fiscal 2026 and 75.0% in Q2 fiscal 2027 23. Because manufacturing is outsourced, capital spending is small relative to revenue (see the history table). In our view, margins depend mainly on the mix of new-generation systems and on supply costs for memory and advanced packaging rather than on factory utilisation.

Competitive position, drivers and risks

Competitive position

NVIDIA names competitors including AMD, Broadcom, Intel, Marvell, Huawei, Cisco and Arista, as well as internal chip teams at large cloud companies 1. Silicon Analytica view: AMD GPUs and custom cloud accelerators are the most direct alternatives in AI compute.

Evidence

Reported: Data Center revenue grew from US$115.2bn to US$193.7bn in fiscal 2026 and networking revenue more than doubled 1. NVIDIA describes CUDA as running on all its GPUs, which gives developers one software target across generations 1.

Growth drivers and conditions

Company view: demand for AI training and inference ("agentic AI") and rack-scale Blackwell and Vera Rubin systems 2. Conditions: customers must keep funding data-center build-outs, and TSMC packaging, high-bandwidth memory and power availability must keep pace.

Cyclicality and downside risks

Reported: US export controls mean NVIDIA says it cannot currently deliver a competitive data-center product for China that receives approval 1. Supply depends on a few manufacturers concentrated in Taiwan and Korea 1. Our view: spending by a small group of hyperscalers creates cyclical risk if their returns on AI investment disappoint; custom accelerators are the main structural threat to pricing.

What to watch next

  1. Data Center revenue split between Hyperscale and AI Clouds, Industrial & Enterprise
  2. Gross margin as Vera Rubin systems ramp
  3. Networking revenue growth versus compute
  4. Any change in China export licensing
  5. Inventory and supply commitments relative to revenue

Peer comparison

Compare with another company

Suggested peers: AMD, Broadcom, Intel, Qualcomm

AMD is the closest product comparison. Broadcom combines custom AI chips with a large software business, so its consolidated margins are not a like-for-like chip comparison. Intel manufactures its own chips, which changes capital intensity. Financials compares latest reported periods and labels mismatched period ends.

Latest analysis on NVIDIA

Markets

The AI spending boom: more computing, or more expensive computing?

TrendForce expects AI server revenue to almost double in 2027 while shipments rise about a third. The gap shows how much of the boom is higher value per system.

14 Oct 2026●7 min●Analysis
Markets

The economics inside an AI rack

Accelerators dominate the value of a representative AI rack. Lower-cost chips still underpin its operation.

13 Oct 2026●3 min read●Analysis
Markets

Two days, two directions: why chip stocks broke as markets digested the midterms

Wall Street rallied on the election result, then turned on the AI trade. Semiconductors took the hardest hit, even though nothing in the industry's reported numbers changed this week.

5 Nov 2026●4 min read●Analysis
Manufacturing

The Race to 1 nm — and the Chip Economy Beyond It

What process-node names mean, which chips use them, and why manufacturing progress, revenue and capacity do not move in lockstep.

13 Oct 2026●12 min read●Analysis

Compare with

NVIDIA vs AMDNVIDIA vs Broadcom

Themes

AI infrastructure demandMemory & packaging

More data on NVIDIA

NVIDIA income statement and balance sheet

Related coverage and calendar

Markets

Two days, two directions: why chip stocks broke as markets digested the midterms

Wall Street rallied on the election result, then turned on the AI trade. Semiconductors took the hardest hit, even though nothing in the industry's reported numbers changed this week.

5 Nov 2026●4 min read●Analysis
Markets

The AI spending boom: more computing, or more expensive computing?

TrendForce expects AI server revenue to almost double in 2027 while shipments rise about a third. The gap shows how much of the boom is higher value per system.

14 Oct 2026●7 min●Analysis
Manufacturing

The Race to 1 nm — and the Chip Economy Beyond It

What process-node names mean, which chips use them, and why manufacturing progress, revenue and capacity do not move in lockstep.

13 Oct 2026●12 min read●Analysis
Sources & methodology
  1. NVIDIA Form 10-K for fiscal 2026 ↗ · 25 Feb 2026 · Item 1 Business; Item 7 MD&A; segment and revenue notes
  2. NVIDIA Q4 and fiscal 2026 results release (Form 8-K) ↗ · 25 Feb 2026 · Press release and CFO commentary
  3. NVIDIA Q2 fiscal 2027 CFO commentary (Form 8-K) ↗ · 26 Aug 2026 · Summary tables, revenue by market platform, outlook

Annual and quarterly figures come from the company's SEC filings (XBRL financial data and, where needed, the filed statements). Each value keeps its filing, period dates and accounting basis. Ratios, growth rates, free cash flow, split-adjusted EPS and derived quarters are Silicon Analytica calculations, shown with their inputs. Percentage growth is not shown when the prior value is zero or negative. Guidance is shown only as company guidance; no analyst consensus is used.

Data checks: ✓ 62 passed · ⚠ R&D change Q2 FY2027: 64.4% YoY exceeds the 60% review threshold.

AW · Reviewed 14 Oct 2026