NVIDIA Reports $115.8 Billion Quarter as Vera Rubin Ships
NVIDIA says revenue more than doubled from a year earlier and Vera Rubin NVL72 racks are now shipping. The outlook assumes no China data centre sales, and the company is led by an interim chief executive.
By Utopia Data & AI Team6 min read
In brief
- NVIDIA reported third-quarter fiscal 2027 revenue of $115.8 billion for the quarter ended 25 October 2026, up 20% from the previous quarter and 103% from a year ago. Data Center revenue was $107.6 billion.[1]
- The company guided to $130.0 billion, plus or minus 3%, for the fourth quarter and said it assumes no Data Center compute revenue from China. Vera Rubin NVL72 has begun shipping to cloud partners.[1]
- NVIDIA also announced new supply agreements and a board committee on national security and export controls. Interim CEO Colette Kress leads the company. All figures are NVIDIA's own and we did not audit them.[1][2][3]
What NVIDIA reported on 17 November
NVIDIA reported revenue of $115.8 billion for its third quarter of fiscal 2027, which ended on 25 October 2026. The figure is up 20% from the previous quarter and up 103% from a year earlier. The company published the results on Tuesday 17 November 2026, and the release says gross margins were 74.8% on a GAAP basis and 74.9% on a non-GAAP basis.[1]
Earnings per diluted share were $2.70 on a GAAP basis and $2.68 non-GAAP. Colette Kress, who is interim chief executive and also the chief financial officer, said in the release that the company's momentum is real and that Vera Rubin is off to an explosive start. The release describes Vera Rubin NVL72 as the fastest product ramp in NVIDIA's history.[1]
This story relies on NVIDIA's own press release and two companion announcements. The numbers are the company's, they come from its own accounts, and we have not audited them or compared them with analyst forecasts.[1][2][3]
Revenue, margins and the fourth-quarter outlook
Data Center revenue reached $107.6 billion, up 21% from the previous quarter and up 110% from a year ago. NVIDIA's separate Edge Computing line, which covers products such as RTX Spark laptops, DGX Spark systems and gaming, was $8.2 billion, up 14% on the quarter and 42% on the year.[1]
For the fourth quarter, which includes an extra week, NVIDIA expects revenue of $130.0 billion, plus or minus 3%. It expects GAAP and non-GAAP gross margins between 71.0% and 72.0%, lower than the third quarter. Operating expenses are expected at about $11.0 billion on a GAAP basis and $10.7 billion non-GAAP.[1]
The outlook states that NVIDIA is not assuming any Data Center compute revenue from China. That matters because it removes one large source of uncertainty from the guidance, but it also means any change in export rules could move results in either direction. The release gives no China revenue figure for the quarter just ended.[1]
Vera Rubin NVL72 moves from announcement to shipments
The release says NVIDIA ramped production and began shipping Vera Rubin NVL72 racks during the quarter. It lists racks running at CoreWeave, Google Cloud, Microsoft Azure, Oracle Cloud Infrastructure and Nebius, and coming online soon at SpaceXSI. It also names general availability on Oracle Cloud Infrastructure and Google Cloud's Vera Rubin-powered A5X instances.[1]
A companion announcement on supply says Vera Rubin NVL72 and Vera CPUs began volume shipments in the third quarter, using TSMC's 5nm and 3nm processes and its CoWoS advanced packaging. NVIDIA says it is working with Amkor, Foxconn, Kinsus, Samsung Electro-Mechanics, SPIL and TSMC to expand capacity for substrates, packaging, wafers and system assembly.[2]
One concrete figure in that announcement concerns Kinsus, which plans to expand its ABF package substrate capacity dedicated to NVIDIA by more than 3.5 times from its Taoyuan campus in Taiwan. Production is expected to begin in the second half of 2027, with the added capacity targeted for the third quarter of 2028.[2]
Customers, Hugging Face and the agent stack
The release lists a wide set of commitments. Meta plans expanded deployment of Vera Rubin, Vera CPUs, Spectrum-X networking and BlueField-4 for its Muse agent and model family. OpenAI, CoreWeave, Nebius, Nscale and Verda are building with Vera CPUs, and NVIDIA says it expanded Anthropic's access to Grace Blackwell and Vera Rubin capacity through cloud providers.[1]
The company also repeated that it has agreed to acquire Hugging Face, and that the deal is pending. It described an Open Agent Safety Platform and an OpenShell reference architecture that uses BlueField-4 DPUs to monitor agent behaviour. NVIDIA reported a planned buildout of up to 2 gigawatts of DSX AI factories in Australia by 2027.[1]
On the edge side, the release says RTX Spark laptops are shipping from ASUS, Dell, HP, Lenovo, Microsoft Surface and MSI, and that a DGX Spark 64GB model is available. These are listed as achievements in the quarter, not as separate revenue figures, so the release alone does not show how much each product earned.[1]
Governance, export controls and shareholder returns
On the same day, NVIDIA's board created a National Security and Export Control Oversight Committee. Director Suzanne Nora Johnson chairs it, with directors Stephen C. Neal and John O. Dabiri as members. It will meet at least quarterly, and the company's national security and global trade compliance teams will report to it and to Kress.[3]
On capital returns, NVIDIA said its board approved a $150 billion increase to its share repurchase authorisation in September 2026. The company returned $28.1 billion to shareholders in the quarter through buybacks and dividends, and about $227.3 billion remains authorised. The next dividend is $0.25 per share, payable on 23 December 2026 to holders of record on 2 December.[1]
Why Malaysia should care
Our analysis: Malaysian data centre operators and cloud buyers compete for the same accelerators, so NVIDIA's supply commitments matter to delivery dates. The release names no Malaysian customer, price or allocation, so ask suppliers for firm lead times.
Malaysian data centre operators
Higher NVIDIA volumes and new substrate and packaging capacity may ease supply, but the capacity targeted for 2027 and 2028 does not help orders placed today.[2]
Practical move: Ask your supplier for a written delivery date and the product generation you will receive.
Malaysian companies buying AI cloud capacity
Vera Rubin racks are running at several large clouds, so rental options may widen. The release gives no regional availability or price.[1]
Practical move: Check whether your cloud region lists Vera Rubin instances and compare ringgit pricing.
What Malaysians can do now
- Read NVIDIA's results release and its supply chain announcement in full.
- Ask your cloud or hardware supplier which Vera Rubin products it can deliver in Malaysia, and when.
- Treat all figures as NVIDIA's own until independent analysis or filings confirm them.
What we still do not know
What we still do not know
- What Vera Rubin NVL72 racks cost, and when they reach Malaysian data centres.
- How much of the quarter's revenue came from each product line, and from China.
- Whether the Hugging Face acquisition closes on the timetable NVIDIA expects.
Sources
- NVIDIA Announces Financial Results for Third Quarter Fiscal 2027NVIDIA Newsroom, 17 November 2026
- NVIDIA Expands Commitments With Supply Chain Partners to Support AI Infrastructure BuildoutNVIDIA Newsroom, 17 November 2026
- NVIDIA Board of Directors Establishes National Security and Export Control Oversight CommitteeNVIDIA Newsroom, 17 November 2026
- File:NVIDIA Headquarters.jpgWikimedia Commons, 4 August 2018
- File:Nvdia Endeavor Grand Opening entry.jpgWikimedia Commons, 12 November 2017
- File:CyrusOne Carrollton - Interior.jpgWikimedia Commons, 12 July 2011
Facts come from NVIDIA Newsroom releases dated 17 November 2026, read on 20 November. Figures and claims are those NVIDIA reported, and we did not audit them. Sections marked as our analysis are Utopia's interpretation. The three photographs are licensed contextual images and none shows the results announcement. No generated image is used.
Last fact check: 20 November 2026. Factual correction? devteam.utopia@gmail.com.


