Nvidia
Nvidia in the enterprise
As of November 5, 2026. Every figure links to its source; company-reported numbers are labeled as such.
Nvidia sells the computing behind most corporate AI. Its revenue comes overwhelmingly from data-center hardware: GPUs, complete AI server racks and networking, bought by cloud giants and, more and more, directly by corporations, AI labs and governments. On top of that hardware it sells software and services: NVIDIA AI Enterprise, NIM ready-to-run model packages, the free Nemotron family of AI models, and an "Agent Toolkit" for building AI agents (software that carries out tasks on its own). These tie customers more closely to its chips. The company does not report software revenue separately.
The scale is unmatched. In the quarter ended Jul 2026 (fiscal Q2 2027), Nvidia reported $96.2 billion in revenue, up 106% from a year earlier, of which $89.0 billion was data center. It guided to $108 billion for the following quarter (company figures). Management said customers other than the big cloud companies (specialist AI clouds, governments, industrial firms and enterprises) produced $40 billion that quarter, up 138% year over year, and now make up close to half of data-center sales. Over the prior 12 months, automakers bought $8 billion of on-premises systems, and financial services, manufacturing and health care together bought $7 billion.
The enterprise story is spreading across industries. Eli Lilly built a more than 1,000-GPU "AI factory" for drug discovery, and Bristol Myers Squibb is following. Trading firms Jane Street and Hudson River Trading run Nvidia-powered AI factories. Business software makers SAP, Salesforce, ServiceNow and Adobe are building agents on Nvidia's toolkit. What to watch: how much revenue still depends on a handful of hyperscalers (about 55% of data-center sales last quarter); and China, where the current forecast assumes no data-center chip sales.
Key numbers
| Measure | Figure | As of | Source |
|---|---|---|---|
| Total revenue | $96.2B (+106% YoY) | Q2 FY2027 (quarter ended Jul 2026) | NVIDIA Announces Financial Results for Second Quarter Fiscal 2027 (NVIDIA Form 8-K, SEC) |
| Data center revenue | $89.0B (+117% YoY) | Q2 FY2027 (quarter ended Jul 2026) | NVIDIA Announces Financial Results for Second Quarter Fiscal 2027 (NVIDIA Form 8-K, SEC) |
| Next-quarter revenue guidance | $108B (±2%), assumes no China data-center compute sales | Aug 2026 | NVIDIA Announces Financial Results for Second Quarter Fiscal 2027 (NVIDIA Form 8-K, SEC) |
| Revenue from non-hyperscale customers (AI clouds, industrial, enterprise, sovereign) | $40B (+138% YoY); hyperscalers $49B | Q2 FY2027 (quarter ended Jul 2026) | NVIDIA Q2 FY2027 Earnings Call Transcript, Aug 26, 2026 (NVIDIA) |
| On-premises AI revenue from automakers | $8B (trailing 12 months) | Aug 2026 | NVIDIA Q2 FY2027 Earnings Call Transcript, Aug 26, 2026 (NVIDIA) |
| On-premises AI revenue from financial services, manufacturing and health care | $7B combined (trailing 12 months) | Aug 2026 | NVIDIA Q2 FY2027 Earnings Call Transcript, Aug 26, 2026 (NVIDIA) |
What it sells to businesses
| Data-center GPUs and rack systems (Blackwell, Vera Rubin) | The AI chips and pre-built server racks that companies and clouds buy to train and run AI models. |
| NVIDIA AI Enterprise | A paid, supported software suite for running AI applications reliably on Nvidia hardware in a company's own data center or cloud. |
| NVIDIA NIM | Pre-packaged AI models that a company can switch on as a ready-made service instead of building the setup itself. |
| DGX SuperPOD / Enterprise AI Factory | Turnkey blueprints and systems for a company to build its own private AI supercomputer. |
| DGX Cloud | Nvidia-managed AI computing rented through partner clouds, for firms that don't want to own hardware. |
| Nemotron models and Agent Toolkit (AI-Q, OpenShell) | Free AI models and building blocks for creating AI agents that act inside business systems, with security guardrails. |
Named customers and deployments
| Customer | What they did | When | Source |
|---|---|---|---|
| Eli Lilly | Pharma "AI factory" supercomputer with more than 1,000 Blackwell Ultra GPUs in Lilly-run facilities, for drug discovery, clinical trials and precision medicine; slated to be operational in Jan 2026. | Oct 2025 | Eli Lilly and Nvidia team up to build an AI supercomputer dedicated to drug discovery (MarketScreener) |
| Bristol Myers Squibb | Investing in a Vera Rubin-based AI factory, following Roche and Lilly buildouts, per Nvidia. | Aug 2026 | NVIDIA Q2 FY2027 Earnings Call Transcript, Aug 26, 2026 (NVIDIA) |
| Jane Street and Hudson River Trading | Using Nvidia-powered AI factories to speed up quantitative trading, per Nvidia. | Aug 2026 | NVIDIA Q2 FY2027 Earnings Call Transcript, Aug 26, 2026 (NVIDIA) |
| Samsung Electronics | Uses Nvidia's cuLitho software for up to 20x faster computational lithography in chipmaking, per Nvidia. | Aug 2026 | NVIDIA Q2 FY2027 Earnings Call Transcript, Aug 26, 2026 (NVIDIA) |
| Apple | Apple's Private Cloud Compute uses Nvidia GPUs with Confidential Computing for private AI processing, per Nvidia. | Aug 2026 | NVIDIA Announces Financial Results for Second Quarter Fiscal 2027 (NVIDIA Form 8-K, SEC) |
| IQVIA | Has deployed more than 150 AI agents, including for 19 of the top 20 pharma companies, and is building Nvidia's Nemotron models into its IQVIA.ai platform. | Mar 2026 | Nvidia Expands AI Agent Ecosystem With New Toolkit for Adobe, SAP, Salesforce (eWeek) |
| SK Telecom, NAVER and Brookfield (South Korea) | Building gigawatt-scale sovereign AI infrastructure on Nvidia's DSX platform, per Nvidia. | Aug 2026 | NVIDIA Announces Financial Results for Second Quarter Fiscal 2027 (NVIDIA Form 8-K, SEC) |
| Amazon Web Services | Deploying an additional 2 million Nvidia GPUs and offering Nvidia's Nemotron models on Amazon Bedrock and SageMaker, per Nvidia. | Aug 2026 | NVIDIA Q2 FY2027 Earnings Call Transcript, Aug 26, 2026 (NVIDIA) |
Partnerships
| Partner | Deal | When | Source |
|---|---|---|---|
| SAP, Salesforce, Adobe, ServiceNow, Atlassian, Siemens, CrowdStrike, Cisco and others | Adopted Nvidia's Agent Toolkit to build AI agents into their business software (e.g., SAP's Joule Studio, Salesforce Agentforce via Slack). | Mar 2026 | Nvidia Expands AI Agent Ecosystem With New Toolkit for Adobe, SAP, Salesforce (eWeek) |
| ServiceNow | Integrating Nvidia's Enterprise AI Factory design with ServiceNow's AI Control Tower so companies can govern and monitor their AI agents. | Mar 2026 | NVIDIA GTC 2026: Governing the Autonomous Workforce (ServiceNow) |
| Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs and KKR | Compute-financing platforms aiming to mobilize over $500B of outside capital to fund AI data centers over time (subject to definitive agreements). | Aug 2026 | NVIDIA Announces Financial Results for Second Quarter Fiscal 2027 (NVIDIA Form 8-K, SEC) |
| CoreWeave, Google Cloud, Microsoft Azure, Oracle Cloud Infrastructure and Nebius | Running Nvidia's new Vera Rubin racks, making them rentable to enterprise customers. | Aug 2026 | NVIDIA Announces Financial Results for Second Quarter Fiscal 2027 (NVIDIA Form 8-K, SEC) |