Data CenterWide moat

Nvidia (NVDA) — moat facet

From $317 million to $193.7 billion in eleven years; a single quarter now outsells every full year Nvidia had before fiscal 2025.

Data Center is Nvidia now. It brought in $193,737 million in fiscal 2026, 89.7% of revenue12, and $89,023 million in the second quarter of fiscal 2027 alone, up 117% from a year earlier3. A single quarter of this line is larger than Nvidia's whole revenue in any fiscal year before 20254.

Data Center revenue growth by fiscal year (%)7%FY16145%FY17133%FY1852%FY192%FY20124%FY2159%FY2241%FY23217%FY24142%FY2568%FY26Calc from NVIDIA Forms 10-K FY2016-FY2026
Eleven years of growth with one flat year, fiscal 2020, when hyperscale buying paused; the compound rate is about 79% a year.

What is inside it has changed as much as its size. The line holds compute, meaning GPUs, the Grace CPU and the systems built around them, and networking, meaning the NVLink, InfiniBand and Ethernet products that join thousands of chips into one machine, plus AI software sold on licence5. In fiscal 2026 compute was $162,361 million and networking $31,376 million; networking grew 142% and compute 59%, driven by the NVLink fabric in the GB200 and GB300 rack systems6. The product itself has moved up from the chip: the annual report describes the business model shifting from Hopper HGX systems to Blackwell full-scale datacenter solutions, a change it also names as a reason gross margin fell7.

The history is short and steep. Nvidia dates the GPU to 1999 and CUDA, the software that opened it to general computing, to 2006; in 2012 the AlexNet network, trained on Nvidia GPUs, won the ImageNet competition8. In fiscal 2015 the datacenter business, then sold under the Tesla brand to researchers working on artificial intelligence and big data, brought in $317 million9. The Mellanox acquisition, closed in April 2020 for about $7 billion, added InfiniBand and high-speed Ethernet10, and fiscal 2021 revenue more than doubled to $6,696 million11. Grace, the first data-center CPU, arrived in 202312. Networking has grown into a business of its own: $3,688 million in fiscal 202313, $31,376 million in fiscal 202614 and $14.8 billion in the first quarter of fiscal 2027 alone, up 199%, against compute's 77%15.

It is paid for as hardware, mostly in large orders. Nvidia's direct customers are server makers, original design manufacturers, distributors, system integrators, cloud providers and AI model makers; the clouds and labs that run the chips often buy through the builders, which is why the concentration numbers need care16. Nvidia does not publish a price per GPU or per rack, so the pricing shows up only in the margins.

Those margins are the best in the company. Compute & Networking, the segment that holds this line and Automotive, earned an operating margin of 67.3% in fiscal 2026, against 71.3% in fiscal 2025 and 33.7% in fiscal 202317. The fiscal 2026 dip came partly from a $4.5 billion charge for H20 inventory and purchase obligations after export rules stranded the China-only chip18.

The growth series is the whole AI boom in twelve numbers: $317 million in fiscal 2015, then $339 million, $830 million, $1,932 million19, $2,932 million and $2,983 million20, $6,696 million, $10,613 million and $15,005 million21, then $47,525 million, $115,186 million and $193,737 million22. That is about 79% a year compounded over eleven years23. The weak year was fiscal 2020, up 2%, which the annual report put down to vertical-industry growth offset by lower hyperscale sales24. The latest two quarters were $75,246 million, up 92%25, and $89,023 million, up 117%, of which Hyperscale was $48,710 million and AI Clouds, Industrial & Enterprise $40,313 million26.

The outlook is guided and committed. Nvidia pointed the third quarter at $108.0 billion of total revenue with no data-center compute from China in the number27, the chief financial officer told investors to expect revenue growth of about 70% in fiscal 202828, and supply commitments rose to $279 billion, mostly memory29. Inventory rose to $31.6 billion, which Nvidia says is preparation for introducing Vera Rubin in the third quarter30. China, which took a fifth of revenue in fiscal 202431, now takes almost no data-center product: Hopper shipments there were less than 1% of Data Center revenue in the second quarter32. The customers behind that demand, and the ones building chips of their own, are argued on Major Clients and Customers Turned Competitors rather than here.

The line has now grown for ten quarters in a row. The figure that would change the verdict is a quarter in which data-center revenue falls without a product transition to explain it; the last dip in sequential growth, to 5% during the Blackwell handover, recovered within a quarter33.

Moat trajectory: Widening

Revenue grew 68% in fiscal 2026 and 117% in the second quarter of fiscal 2027, networking grew 142% in the year, and the third quarter is guided to $108.0 billion of total revenue without China. The segment margin eased from 71.3% to 67.3%, largely on a $4.5 billion H20 charge, not on price.

The number that tests this moat
Moat Explorer calc
Data Center sequential growth, latest quarter
+18.3% in Q2 FY2027 ($89.0B from $75.2B)

The line has grown for ten quarters in a row; the last dip, to 5.1% during the Blackwell handover, recovered within a quarter. A fall without a product transition to explain it would be the first sign the build-out, not a rival, had peaked.

How it's calculated: Data Center $89,023M (Q2 FY2027) / $75,246M (Q1 FY2027) - 1 = 18.3%.
Source: Moat Explorer calc from NVIDIA Q1 and Q2 FY2027 CFO commentaries ↗
References
  1. ReportedIt brought in $193,737 million in fiscal 2026, 89.7% of revenue, and $89,023 million in the second quarter of fiscal 2027 alone, up 117% from a year earlier.
    NVIDIA Form 10-K, FY2026 - revenue by end market: Data Center $193,737M (compute $162,361M, networking $31,376M), Gaming $16,042M, Professional Visualization $3,191M, Automotive $2,349M, OEM and Other $619M, total $215,938M, with FY2025 ($115,186M; $11,350M; $1,878M; $1,694M; $389M) and FY2024 ($47,525M; $10,447M; $1,553M; $1,091M; $306M); segment operating income Compute & Networking $130,141M on $193,479M and Graphics $9,156M on $22,459M (FY2025 $82,875M on $116,193M and $5,085M on $14,304M); segment definitions; FY2026 growth by market; supply constraints expected to be a headwind to Gaming; the H20 charge; history (GPU 1999, CUDA 2006, AlexNet 2012, Grace 2023); gaming, workstation and DGX Spark product descriptions — FY2024-FY2026 · publ. Filed Feb 2026 · source ↗
  2. Moat Explorer calcIt brought in $193,737 million in fiscal 2026, 89.7% of revenue, and $89,023 million in the second quarter of fiscal 2027 alone, up 117% from a year earlier.
    Moat Explorer calculations from NVIDIA Forms 10-K FY2016-FY2026, Forms 10-Q and CFO commentaries: shares of revenue, compound growth rates FY2015-FY2026, segment margins, fourth-quarter figures (annual less nine months) and unallocated costs (FY2026 operating income $130,387M less segment total $139,297M) — FY2015-Q2 FY2027 · publ. 2026-09-18 · source ↗
  3. ReportedIt brought in $193,737 million in fiscal 2026, 89.7% of revenue, and $89,023 million in the second quarter of fiscal 2027 alone, up 117% from a year earlier.
    NVIDIA Q2 FY2027 CFO commentary (Form 8-K, exhibit 99.2) — Data Center revenue of $89,023M splits into Hyperscale $48,710M (+102% year on year, +13% sequentially), AI Clouds/Industrial & Enterprise $40,313M (+138%, +25%) and Edge Computing $7,198M (+27%, +13%); shipments of Data Center Hopper products to China during the quarter were less than 1% of Data Center revenue; purchase commitments increased from $119 billion in the prior quarter to $279 billion, primarily related to the procurement of memory; segment revenue was Compute & Networking $88,299M (+114%) and Graphics $7,922M (+46%) — Q2 FY2027 (quarter ended 26 July 2026) · publ. 2026-08-26 · source ↗
  4. Moat Explorer calcA single quarter of this line is larger than Nvidia's whole revenue in any fiscal year before 2025.
    Moat Explorer calculations from NVIDIA Forms 10-K FY2016-FY2026, Forms 10-Q and CFO commentaries: shares of revenue, compound growth rates FY2015-FY2026, segment margins, fourth-quarter figures (annual less nine months) and unallocated costs (FY2026 operating income $130,387M less segment total $139,297M) — FY2015-Q2 FY2027 · publ. 2026-09-18 · source ↗
  5. ReportedThe line holds compute, meaning GPUs, the Grace CPU and the systems built around them, and networking, meaning the NVLink, InfiniBand and Ethernet products that join thousands of chips into one machine, plus AI software sold on licence.
    NVIDIA Form 10-K, FY2026 - revenue by end market: Data Center $193,737M (compute $162,361M, networking $31,376M), Gaming $16,042M, Professional Visualization $3,191M, Automotive $2,349M, OEM and Other $619M, total $215,938M, with FY2025 ($115,186M; $11,350M; $1,878M; $1,694M; $389M) and FY2024 ($47,525M; $10,447M; $1,553M; $1,091M; $306M); segment operating income Compute & Networking $130,141M on $193,479M and Graphics $9,156M on $22,459M (FY2025 $82,875M on $116,193M and $5,085M on $14,304M); segment definitions; FY2026 growth by market; supply constraints expected to be a headwind to Gaming; the H20 charge; history (GPU 1999, CUDA 2006, AlexNet 2012, Grace 2023); gaming, workstation and DGX Spark product descriptions — FY2024-FY2026 · publ. Filed Feb 2026 · source ↗
  6. ReportedIn fiscal 2026 compute was $162,361 million and networking $31,376 million; networking grew 142% and compute 59%, driven by the NVLink fabric in the GB200 and GB300 rack systems.
    NVIDIA Form 10-K, FY2026 - revenue by end market: Data Center $193,737M (compute $162,361M, networking $31,376M), Gaming $16,042M, Professional Visualization $3,191M, Automotive $2,349M, OEM and Other $619M, total $215,938M, with FY2025 ($115,186M; $11,350M; $1,878M; $1,694M; $389M) and FY2024 ($47,525M; $10,447M; $1,553M; $1,091M; $306M); segment operating income Compute & Networking $130,141M on $193,479M and Graphics $9,156M on $22,459M (FY2025 $82,875M on $116,193M and $5,085M on $14,304M); segment definitions; FY2026 growth by market; supply constraints expected to be a headwind to Gaming; the H20 charge; history (GPU 1999, CUDA 2006, AlexNet 2012, Grace 2023); gaming, workstation and DGX Spark product descriptions — FY2024-FY2026 · publ. Filed Feb 2026 · source ↗
  7. ReportedThe product itself has moved up from the chip: the annual report describes the business model shifting from Hopper HGX systems to Blackwell full-scale datacenter solutions, a change it also names as a reason gross margin fell.
    NVIDIA Form 10-K, FY2026 - revenue by end market: Data Center $193,737M (compute $162,361M, networking $31,376M), Gaming $16,042M, Professional Visualization $3,191M, Automotive $2,349M, OEM and Other $619M, total $215,938M, with FY2025 ($115,186M; $11,350M; $1,878M; $1,694M; $389M) and FY2024 ($47,525M; $10,447M; $1,553M; $1,091M; $306M); segment operating income Compute & Networking $130,141M on $193,479M and Graphics $9,156M on $22,459M (FY2025 $82,875M on $116,193M and $5,085M on $14,304M); segment definitions; FY2026 growth by market; supply constraints expected to be a headwind to Gaming; the H20 charge; history (GPU 1999, CUDA 2006, AlexNet 2012, Grace 2023); gaming, workstation and DGX Spark product descriptions — FY2024-FY2026 · publ. Filed Feb 2026 · source ↗
  8. ReportedNvidia dates the GPU to 1999 and CUDA, the software that opened it to general computing, to 2006; in 2012 the AlexNet network, trained on Nvidia GPUs, won the ImageNet competition.
    NVIDIA Form 10-K, FY2026 - revenue by end market: Data Center $193,737M (compute $162,361M, networking $31,376M), Gaming $16,042M, Professional Visualization $3,191M, Automotive $2,349M, OEM and Other $619M, total $215,938M, with FY2025 ($115,186M; $11,350M; $1,878M; $1,694M; $389M) and FY2024 ($47,525M; $10,447M; $1,553M; $1,091M; $306M); segment operating income Compute & Networking $130,141M on $193,479M and Graphics $9,156M on $22,459M (FY2025 $82,875M on $116,193M and $5,085M on $14,304M); segment definitions; FY2026 growth by market; supply constraints expected to be a headwind to Gaming; the H20 charge; history (GPU 1999, CUDA 2006, AlexNet 2012, Grace 2023); gaming, workstation and DGX Spark product descriptions — FY2024-FY2026 · publ. Filed Feb 2026 · source ↗
  9. ReportedIn fiscal 2015 the datacenter business, then sold under the Tesla brand to researchers working on artificial intelligence and big data, brought in $317 million.
    NVIDIA Form 10-K, FY2016 - revenue by specialized markets FY2014-FY2016 (FY2015: Gaming $2,058M, Professional Visualization $795M, Datacenter $317M, Automotive $183M, OEM & IP $1,329M, total $4,682M); Tesla for AI and big-data researchers; an end-to-end self-driving solution of DRIVE PX and DIGITS; Professional Visualization served by working closely with independent software vendors; products in more than 10 million cars; GeForce Experience downloaded by more than 70 million users; Intel's 2011 patent cross-licence paying $1.50 billion over six years, final $200 million received January 2016, recognised through Q1 FY2018 — FY2014-FY2016 · publ. Filed Mar 2016 · source ↗
  10. ReportedThe Mellanox acquisition, closed in April 2020 for about $7 billion, added InfiniBand and high-speed Ethernet, and fiscal 2021 revenue more than doubled to $6,696 million.
    NVIDIA — Mellanox Technologies acquisition completed (≈$7.0B; InfiniBand/Ethernet networking) — Closed Apr 2020 · publ. Apr 2020 · source ↗
  11. ReportedThe Mellanox acquisition, closed in April 2020 for about $7 billion, added InfiniBand and high-speed Ethernet, and fiscal 2021 revenue more than doubled to $6,696 million.
    NVIDIA Form 10-K, FY2023 - revenue by specialized markets FY2021-FY2023 (Data Center $6,696M / $10,613M / $15,005M; Gaming $7,759M / $12,462M / $9,067M; Professional Visualization $1,053M / $2,111M / $1,544M; Automotive $536M / $566M / $903M; OEM & Other $631M / $1,162M / $455M) and FY2023 drivers: Gaming and Professional Visualization down 27% on lower sell-in to reduce channel inventory, Automotive up 60%, OEM and Other down 61%; CMP revenue nominal in FY2023 and $550 million in FY2022, provided to address demand from gamers and direct miners to CMP — FY2021-FY2023 · publ. Filed Feb 2023 · source ↗
  12. ReportedGrace, the first data-center CPU, arrived in 2023.
    NVIDIA Form 10-K, FY2026 - revenue by end market: Data Center $193,737M (compute $162,361M, networking $31,376M), Gaming $16,042M, Professional Visualization $3,191M, Automotive $2,349M, OEM and Other $619M, total $215,938M, with FY2025 ($115,186M; $11,350M; $1,878M; $1,694M; $389M) and FY2024 ($47,525M; $10,447M; $1,553M; $1,091M; $306M); segment operating income Compute & Networking $130,141M on $193,479M and Graphics $9,156M on $22,459M (FY2025 $82,875M on $116,193M and $5,085M on $14,304M); segment definitions; FY2026 growth by market; supply constraints expected to be a headwind to Gaming; the H20 charge; history (GPU 1999, CUDA 2006, AlexNet 2012, Grace 2023); gaming, workstation and DGX Spark product descriptions — FY2024-FY2026 · publ. Filed Feb 2026 · source ↗
  13. ReportedNetworking has grown into a business of its own: $3,688 million in fiscal 2023, $31,376 million in fiscal 2026 and $14.8 billion in the first quarter of fiscal 2027 alone, up 199%, against compute's 77%.
    NVIDIA Form 10-K, FY2025 - revenue by end market FY2023-FY2025, segment operating income FY2023-FY2025 (Compute & Networking $5,083M on $15,068M in FY2023), networking $3,688M in FY2023; Graphics revenue $11,906M in FY2023; and FY2025 drivers: Automotive up 55% on self-driving platforms, Professional Visualization up 21% on Ada RTX workstations for generative AI design, simulation and engineering — FY2023-FY2025 · publ. Filed Feb 2025 · source ↗
  14. ReportedNetworking has grown into a business of its own: $3,688 million in fiscal 2023, $31,376 million in fiscal 2026 and $14.8 billion in the first quarter of fiscal 2027 alone, up 199%, against compute's 77%.
    NVIDIA Form 10-K, FY2026 - revenue by end market: Data Center $193,737M (compute $162,361M, networking $31,376M), Gaming $16,042M, Professional Visualization $3,191M, Automotive $2,349M, OEM and Other $619M, total $215,938M, with FY2025 ($115,186M; $11,350M; $1,878M; $1,694M; $389M) and FY2024 ($47,525M; $10,447M; $1,553M; $1,091M; $306M); segment operating income Compute & Networking $130,141M on $193,479M and Graphics $9,156M on $22,459M (FY2025 $82,875M on $116,193M and $5,085M on $14,304M); segment definitions; FY2026 growth by market; supply constraints expected to be a headwind to Gaming; the H20 charge; history (GPU 1999, CUDA 2006, AlexNet 2012, Grace 2023); gaming, workstation and DGX Spark product descriptions — FY2024-FY2026 · publ. Filed Feb 2026 · source ↗
  15. ReportedNetworking has grown into a business of its own: $3,688 million in fiscal 2023, $31,376 million in fiscal 2026 and $14.8 billion in the first quarter of fiscal 2027 alone, up 199%, against compute's 77%.
    NVIDIA Q1 FY2027 CFO commentary - new reporting framework: Data Center (Hyperscale; AI Clouds, Industrial & Enterprise) and Edge Computing (PCs, game consoles, workstations, AI-RAN base stations, robotics and automotive); Data Center $75,246M, up 92%, with compute $60.4B (+77%) and networking $14.8B (+199%); Edge Computing $6,369M, up 29% — Q1 FY2027 · publ. May 2026 · source ↗
  16. ReportedNvidia's direct customers are server makers, original design manufacturers, distributors, system integrators, cloud providers and AI model makers; the clouds and labs that run the chips often buy through the builders, which is why the concentration numbers need care.
    NVIDIA Form 10-Q, Q1 FY2027 - direct customers (AIBs, distributors, ODMs, OEMs, CSPs, AI model makers and system integrators) and indirect customers who purchase through them — Q1 FY2027 · publ. May 2026 · source ↗
  17. Moat Explorer calcCompute & Networking, the segment that holds this line and Automotive, earned an operating margin of 67.3% in fiscal 2026, against 71.3% in fiscal 2025 and 33.7% in fiscal 2023.
    Moat Explorer calculations from NVIDIA Forms 10-K FY2016-FY2026, Forms 10-Q and CFO commentaries: shares of revenue, compound growth rates FY2015-FY2026, segment margins, fourth-quarter figures (annual less nine months) and unallocated costs (FY2026 operating income $130,387M less segment total $139,297M) — FY2015-Q2 FY2027 · publ. 2026-09-18 · source ↗
  18. ReportedThe fiscal 2026 dip came partly from a $4.5 billion charge for H20 inventory and purchase obligations after export rules stranded the China-only chip.
    NVIDIA Form 10-K, FY2026 - revenue by end market: Data Center $193,737M (compute $162,361M, networking $31,376M), Gaming $16,042M, Professional Visualization $3,191M, Automotive $2,349M, OEM and Other $619M, total $215,938M, with FY2025 ($115,186M; $11,350M; $1,878M; $1,694M; $389M) and FY2024 ($47,525M; $10,447M; $1,553M; $1,091M; $306M); segment operating income Compute & Networking $130,141M on $193,479M and Graphics $9,156M on $22,459M (FY2025 $82,875M on $116,193M and $5,085M on $14,304M); segment definitions; FY2026 growth by market; supply constraints expected to be a headwind to Gaming; the H20 charge; history (GPU 1999, CUDA 2006, AlexNet 2012, Grace 2023); gaming, workstation and DGX Spark product descriptions — FY2024-FY2026 · publ. Filed Feb 2026 · source ↗
  19. ReportedThe growth series is the whole AI boom in twelve numbers: $317 million in fiscal 2015, then $339 million, $830 million, $1,932 million, $2,932 million and $2,983 million, $6,696 million, $10,613 million and $15,005 million, then $47,525 million, $115,186 million and $193,737 million.
    NVIDIA Form 10-K, FY2018 - products include cryptocurrency-specific GPUs; revenue by specialized markets FY2016-FY2018 (Gaming $2,818M / $4,060M / $5,513M; Professional Visualization $750M / $835M / $934M; Datacenter $339M / $830M / $1,932M; Automotive $320M / $487M / $558M; OEM & IP $783M / $698M / $777M) — FY2016-FY2018 · publ. Filed Feb 2018 · source ↗
  20. ReportedThe growth series is the whole AI boom in twelve numbers: $317 million in fiscal 2015, then $339 million, $830 million, $1,932 million, $2,932 million and $2,983 million, $6,696 million, $10,613 million and $15,005 million, then $47,525 million, $115,186 million and $193,737 million.
    NVIDIA Form 10-K, FY2020 - revenue by specialized markets FY2018-FY2020 (Gaming $5,513M / $6,246M / $5,518M; Professional Visualization $934M / $1,130M / $1,212M; Data Center $1,932M / $2,932M / $2,983M; Automotive $558M / $641M / $700M; OEM & Other $777M / $767M / $505M); Gaming down 12% on lower desktop GPU and gaming-platform SoC sales; Data Center up 2%, vertical industry growth partially offset by lower hyperscale sales; OEM and Other down 34% primarily due to the absence of cryptocurrency-specific product sales; Professional Visualization up 7% on desktop and notebook workstations; Automotive up 9% on AI cockpit solutions and development services agreements — FY2018-FY2020 · publ. Filed Feb 2020 · source ↗
  21. ReportedThe growth series is the whole AI boom in twelve numbers: $317 million in fiscal 2015, then $339 million, $830 million, $1,932 million, $2,932 million and $2,983 million, $6,696 million, $10,613 million and $15,005 million, then $47,525 million, $115,186 million and $193,737 million.
    NVIDIA Form 10-K, FY2023 - revenue by specialized markets FY2021-FY2023 (Data Center $6,696M / $10,613M / $15,005M; Gaming $7,759M / $12,462M / $9,067M; Professional Visualization $1,053M / $2,111M / $1,544M; Automotive $536M / $566M / $903M; OEM & Other $631M / $1,162M / $455M) and FY2023 drivers: Gaming and Professional Visualization down 27% on lower sell-in to reduce channel inventory, Automotive up 60%, OEM and Other down 61%; CMP revenue nominal in FY2023 and $550 million in FY2022, provided to address demand from gamers and direct miners to CMP — FY2021-FY2023 · publ. Filed Feb 2023 · source ↗
  22. ReportedThe growth series is the whole AI boom in twelve numbers: $317 million in fiscal 2015, then $339 million, $830 million, $1,932 million, $2,932 million and $2,983 million, $6,696 million, $10,613 million and $15,005 million, then $47,525 million, $115,186 million and $193,737 million.
    NVIDIA Form 10-K, FY2026 - revenue by end market: Data Center $193,737M (compute $162,361M, networking $31,376M), Gaming $16,042M, Professional Visualization $3,191M, Automotive $2,349M, OEM and Other $619M, total $215,938M, with FY2025 ($115,186M; $11,350M; $1,878M; $1,694M; $389M) and FY2024 ($47,525M; $10,447M; $1,553M; $1,091M; $306M); segment operating income Compute & Networking $130,141M on $193,479M and Graphics $9,156M on $22,459M (FY2025 $82,875M on $116,193M and $5,085M on $14,304M); segment definitions; FY2026 growth by market; supply constraints expected to be a headwind to Gaming; the H20 charge; history (GPU 1999, CUDA 2006, AlexNet 2012, Grace 2023); gaming, workstation and DGX Spark product descriptions — FY2024-FY2026 · publ. Filed Feb 2026 · source ↗
  23. Moat Explorer calcThat is about 79% a year compounded over eleven years.
    Moat Explorer calculations from NVIDIA Forms 10-K FY2016-FY2026, Forms 10-Q and CFO commentaries: shares of revenue, compound growth rates FY2015-FY2026, segment margins, fourth-quarter figures (annual less nine months) and unallocated costs (FY2026 operating income $130,387M less segment total $139,297M) — FY2015-Q2 FY2027 · publ. 2026-09-18 · source ↗
  24. ReportedThe weak year was fiscal 2020, up 2%, which the annual report put down to vertical-industry growth offset by lower hyperscale sales.
    NVIDIA Form 10-K, FY2020 - revenue by specialized markets FY2018-FY2020 (Gaming $5,513M / $6,246M / $5,518M; Professional Visualization $934M / $1,130M / $1,212M; Data Center $1,932M / $2,932M / $2,983M; Automotive $558M / $641M / $700M; OEM & Other $777M / $767M / $505M); Gaming down 12% on lower desktop GPU and gaming-platform SoC sales; Data Center up 2%, vertical industry growth partially offset by lower hyperscale sales; OEM and Other down 34% primarily due to the absence of cryptocurrency-specific product sales; Professional Visualization up 7% on desktop and notebook workstations; Automotive up 9% on AI cockpit solutions and development services agreements — FY2018-FY2020 · publ. Filed Feb 2020 · source ↗
  25. ReportedThe latest two quarters were $75,246 million, up 92%, and $89,023 million, up 117%, of which Hyperscale was $48,710 million and AI Clouds, Industrial & Enterprise $40,313 million.
    NVIDIA Q1 FY2027 CFO commentary - new reporting framework: Data Center (Hyperscale; AI Clouds, Industrial & Enterprise) and Edge Computing (PCs, game consoles, workstations, AI-RAN base stations, robotics and automotive); Data Center $75,246M, up 92%, with compute $60.4B (+77%) and networking $14.8B (+199%); Edge Computing $6,369M, up 29% — Q1 FY2027 · publ. May 2026 · source ↗
  26. ReportedThe latest two quarters were $75,246 million, up 92%, and $89,023 million, up 117%, of which Hyperscale was $48,710 million and AI Clouds, Industrial & Enterprise $40,313 million.
    NVIDIA Q2 FY2027 CFO commentary (Form 8-K, exhibit 99.2) — Data Center revenue of $89,023M splits into Hyperscale $48,710M (+102% year on year, +13% sequentially), AI Clouds/Industrial & Enterprise $40,313M (+138%, +25%) and Edge Computing $7,198M (+27%, +13%); shipments of Data Center Hopper products to China during the quarter were less than 1% of Data Center revenue; purchase commitments increased from $119 billion in the prior quarter to $279 billion, primarily related to the procurement of memory; segment revenue was Compute & Networking $88,299M (+114%) and Graphics $7,922M (+46%) — Q2 FY2027 (quarter ended 26 July 2026) · publ. 2026-08-26 · source ↗
  27. ReportedNvidia pointed the third quarter at $108.0 billion of total revenue with no data-center compute from China in the number, the chief financial officer told investors to expect revenue growth of about 70% in fiscal 2028, and supply commitments rose to $279 billion, mostly memory.
    NVIDIA, Q2 FY2027 financial results (Form 8-K, exhibit 99.1) — revenue of $96,221M for the quarter ended 26 July 2026, up 18% sequentially and 106% year on year; Data Center revenue of $89.0 billion, up 18% sequentially and 117% year on year; GAAP and non-GAAP gross margins both 75.0%, against 74.9% in Q1 FY2027 and 72.4% a year earlier; operating expenses $8,408M; operating income $63,734M, up 124%; GAAP net income $59,688M, up 126%; GAAP diluted EPS $2.46 against non-GAAP $2.22, the two having crossed because from Q1 FY2027 NVIDIA's non-GAAP measures no longer exclude stock-based compensation expense; approximately $26.0 billion returned to shareholders in the quarter with roughly $99.0 billion remaining under the repurchase authorisation; six-month FY2027 revenue $177,837M and net income $118,010M. Outlook for Q3 FY2027: revenue of $108.0 billion plus or minus 2%, with NVIDIA not assuming any Data Center compute revenue from China; GAAP and non-GAAP gross margins of 74.0% plus or minus 50 basis points. Jensen Huang: 'AI has reached its inflection point... Vera Rubin, now in full production, was built to power exactly this moment.' — Q2 FY2027 (quarter ended 26 July 2026) · publ. 2026-08-26 · source ↗
  28. ReportedNvidia pointed the third quarter at $108.0 billion of total revenue with no data-center compute from China in the number, the chief financial officer told investors to expect revenue growth of about 70% in fiscal 2028, and supply commitments rose to $279 billion, mostly memory.
    NVIDIA Q2 FY2027 earnings call, reported by CNBC — chief financial officer Colette Kress told investors that revenue growth in fiscal 2028 would reach about 70%, against an average analyst estimate nearer 44%; on a consensus fiscal 2027 revenue projection of roughly $396 billion that implies fiscal 2028 revenue of about $673 billion — FY2028 outlook · publ. 2026-08-26 · source ↗
  29. ReportedNvidia pointed the third quarter at $108.0 billion of total revenue with no data-center compute from China in the number, the chief financial officer told investors to expect revenue growth of about 70% in fiscal 2028, and supply commitments rose to $279 billion, mostly memory.
    NVIDIA Q2 FY2027 CFO commentary (Form 8-K, exhibit 99.2) — Data Center revenue of $89,023M splits into Hyperscale $48,710M (+102% year on year, +13% sequentially), AI Clouds/Industrial & Enterprise $40,313M (+138%, +25%) and Edge Computing $7,198M (+27%, +13%); shipments of Data Center Hopper products to China during the quarter were less than 1% of Data Center revenue; purchase commitments increased from $119 billion in the prior quarter to $279 billion, primarily related to the procurement of memory; segment revenue was Compute & Networking $88,299M (+114%) and Graphics $7,922M (+46%) — Q2 FY2027 (quarter ended 26 July 2026) · publ. 2026-08-26 · source ↗
  30. ReportedInventory rose to $31.6 billion, which Nvidia says is preparation for introducing Vera Rubin in the third quarter.
    NVIDIA Q2 FY2027 CFO commentary (Form 8-K, exhibit 99.2) — Data Center revenue of $89,023M splits into Hyperscale $48,710M (+102% year on year, +13% sequentially), AI Clouds/Industrial & Enterprise $40,313M (+138%, +25%) and Edge Computing $7,198M (+27%, +13%); shipments of Data Center Hopper products to China during the quarter were less than 1% of Data Center revenue; purchase commitments increased from $119 billion in the prior quarter to $279 billion, primarily related to the procurement of memory; segment revenue was Compute & Networking $88,299M (+114%) and Graphics $7,922M (+46%) — Q2 FY2027 (quarter ended 26 July 2026) · publ. 2026-08-26 · source ↗
  31. ReportedChina, which took a fifth of revenue in fiscal 2024, now takes almost no data-center product: Hopper shipments there were less than 1% of Data Center revenue in the second quarter.
    NVIDIA Form 10-K, FY2026 - revenue by end market: Data Center $193,737M (compute $162,361M, networking $31,376M), Gaming $16,042M, Professional Visualization $3,191M, Automotive $2,349M, OEM and Other $619M, total $215,938M, with FY2025 ($115,186M; $11,350M; $1,878M; $1,694M; $389M) and FY2024 ($47,525M; $10,447M; $1,553M; $1,091M; $306M); segment operating income Compute & Networking $130,141M on $193,479M and Graphics $9,156M on $22,459M (FY2025 $82,875M on $116,193M and $5,085M on $14,304M); segment definitions; FY2026 growth by market; supply constraints expected to be a headwind to Gaming; the H20 charge; history (GPU 1999, CUDA 2006, AlexNet 2012, Grace 2023); gaming, workstation and DGX Spark product descriptions — FY2024-FY2026 · publ. Filed Feb 2026 · source ↗
  32. ReportedChina, which took a fifth of revenue in fiscal 2024, now takes almost no data-center product: Hopper shipments there were less than 1% of Data Center revenue in the second quarter.
    NVIDIA Q2 FY2027 CFO commentary (Form 8-K, exhibit 99.2) — Data Center revenue of $89,023M splits into Hyperscale $48,710M (+102% year on year, +13% sequentially), AI Clouds/Industrial & Enterprise $40,313M (+138%, +25%) and Edge Computing $7,198M (+27%, +13%); shipments of Data Center Hopper products to China during the quarter were less than 1% of Data Center revenue; purchase commitments increased from $119 billion in the prior quarter to $279 billion, primarily related to the procurement of memory; segment revenue was Compute & Networking $88,299M (+114%) and Graphics $7,922M (+46%) — Q2 FY2027 (quarter ended 26 July 2026) · publ. 2026-08-26 · source ↗
  33. Moat Explorer calcThe figure that would change the verdict is a quarter in which data-center revenue falls without a product transition to explain it; the last dip in sequential growth, to 5% during the Blackwell handover, recovered within a quarter.
    Moat Explorer calculations from NVIDIA Forms 10-K FY2016-FY2026, Forms 10-Q and CFO commentaries: shares of revenue, compound growth rates FY2015-FY2026, segment margins, fourth-quarter figures (annual less nine months) and unallocated costs (FY2026 operating income $130,387M less segment total $139,297M) — FY2015-Q2 FY2027 · publ. 2026-09-18 · source ↗
Sources
Generated September 18, 2026