Nvidia’s Operating Profit Jumps 115% as Rubin Revenue Nears $20 Billion
IT DAILY · · 1 views
✦ AI Summary
Nvidia announced fiscal 2027 third-quarter revenue of USD 115.844 billion, GAAP operating profit of USD 77.52 billion, and net income of USD 65.287 billion.
Data center revenue was USD 107.61 billion, or about 93% of total revenue, and Nvidia said Vera Rubin generated revenue approaching USD 20 billion in its first quarter of production shipments.
Nvidia guided fiscal 2027 fourth-quarter revenue at USD 130 billion, plus or minus 3%, and gross margin at 71% to 72%, while raising its fiscal 2028 revenue growth outlook from about 70% to 75%.
BizCrushWant to follow Korean talks and live streams in real time?Try CapCrush on the web
Nvidia reported its fiscal 2027 third-quarter results on the 17th, citing expanded supply of its next-generation AI semiconductor platform, Vera Rubin. The reporting period was July 27 to October 25, 2026. GAAP operating profit for the period came to USD 77.52 billion, up 115% from a year earlier and more than doubling. Net income rose 105% from a year earlier to USD 65.287 billion, while revenue increased 103% to USD 115.844 billion. Gross margin rose 1.4 percentage points to 74.8%.
Nvidia is accelerating efforts to secure manufacturing capacity by expanding long-term supply agreements centered on memory procurement and signing a new agreement with Samsung Electro-Mechanics in substrates. The company is pursuing expanded long-term supply agreements centered on memory procurement and a new substrate agreement with Samsung Electro-Mechanics to secure manufacturing capacity.
Against this backdrop, Nvidia raised its forecast for fiscal 2028 revenue growth.
Data center revenue came to USD 107.61 billion, up 110% from a year earlier. The data center segment accounted for about 93% of total revenue.
Within that, the hyperscale segment targeted large cloud providers and posted revenue of USD 60.136 billion, up 71%. Revenue from the AI cloud, industrial and enterprise segment rose 195% to USD 47.474 billion.
These figures were linked to a trend in which demand for AI infrastructure is expanding to specialized cloud and enterprise customers. The next-generation platform was adopted by major data centers within three months of shipment.
A notable change in this earnings release was the rapid revenue contribution from Vera Rubin. Colette Kress, Nvidia's interim CEO and CFO, said Vera Rubin generated revenue approaching USD 20 billion in its first quarter of production shipments. She described this as the fastest pace of new product ramp-up in Nvidia's history.
After starting Vera Rubin production shipments in August, Nvidia supplied Vera Rubin to data centers operated by major hyperscale and AI-specialized cloud providers in about three months. The time it took major hyperscale customers to begin using Vera Rubin in actual operations was about one month.
That adoption period was about half that of the existing Grace Blackwell. As demand for the existing Blackwell products continued, Vera Rubin supply moved into full swing.
This trend lifted the company’s forward guidance. Nvidia guided for fiscal 2027 fourth-quarter revenue of USD 130 billion, plus or minus 3%. The quarter is typically one week longer than a normal quarter. Revenue from computing products for China data centers was not included in the forecast.
Nvidia raised its fiscal 2028 revenue growth outlook to about 70% to 75%. Interim CEO Kress said rising customer demand and additional supply capacity secured the higher outlook. Kress also said the revision reflected customers' readiness to build out data centers and their ability to secure financing.
Nvidia is sticking to its next-generation product roadmap. The planned unveiling of Vera Rubin Ultra is set for 2027, and the planned unveiling of the follow-on architecture, Feynman, is set for 2028.
Nvidia is moving to expand its supply chain in earnest to respond to surging demand for AI semiconductors. To that end, it is reviewing support for capacity expansions in wafers, substrates and advanced packaging.
According to an explanation document from Nvidia's CFO, the scale of supply and production-capacity commitments at the end of the third quarter was USD 390 billion. The comparable figure in the previous quarter was USD 279 billion.
The increase in commitments was USD 111 billion. The company said the main driver of the increase was multi-year memory procurement.
The USD 390 billion figure includes total supply commitments covering memory, other components and production capacity. The core of the increase is long-term memory procurement.
This is drawing attention to Nvidia's move to preemptively secure memory supply for next-generation AI semiconductor production. Nvidia is also expanding the scope of cooperation into substrates and advanced packaging.
Nvidia is broadening supply chain cooperation in response to a situation in which customer demand exceeds supply capacity. Interim CEO Kress said the company signed new substrate agreements with Samsung Electro-Mechanics and Kinsus of Taiwan, and Nvidia said it is also working with Taiwan's Siliconware Precision Industries (SPIL) in advanced packaging. This cooperation follows supply-chain expansion measures announced in October with Amkor Technology, KYEC and Unimicron.
Nvidia is pursuing expanded production capacity for key components such as wafers, substrates and advanced packaging. However, it did not disclose specific items covered by the agreement with Samsung Electro-Mechanics, and it did not reveal the contract value.
The company is also considering financial support to help suppliers expand production capacity. Interim CEO Kress said supporting suppliers is a capital-allocation priority, and added that the company could support capacity expansion through prepayments if needed.
Nvidia said actual data center operating conditions are becoming as important as semiconductor production capacity. It said that when allocating products to customers, it also considers the readiness of data center sites, power, buildings and financing capacity, adding that securing semiconductors alone is not enough and that installed, operational facilities and funding are also needed.
Nvidia pointed to costs tied to supply expansion and the continued U.S. export restrictions on China as variables in future results. The company expects profitability in fiscal 2027 fourth quarter to be lower than in the third quarter because of higher costs during the supply expansion process and the impact of existing pricing terms.
Nvidia's fiscal 2027 fourth-quarter gross margin outlook is 71% to 72%. That is below the third-quarter gross margin of 74.8%.
Interim CEO Kress said the fourth-quarter gross margin outlook was affected by the reflection of higher supply costs and pricing terms applied to existing customer orders. In the company's view, fourth-quarter gross margin marks the low point.
Nvidia plans to switch to new pricing terms afterward. As it pursues a recovery in profitability through the transition to those pricing terms, Nvidia expects fiscal 2028 gross margin to be 72% to 73%.
At the same time, uncertainty in its China business continues. Revenue from Hopper data center products eligible for China export licenses was USD 2.4 billion in the third quarter, but Nvidia said it has now halted shipments of those products to China. As a result, Nvidia did not include revenue from computing products for China data centers in its fourth-quarter outlook.
This earnings release was the first official event held since CEO Jensen Huang stepped down on the 12th. Nvidia's board named CFO Kress as interim CEO while also beginning the process of selecting a successor.
Interim CEO Kress said the management change will not alter the existing business or supply-chain cooperation. The company has secured memory supply capacity and advanced packaging supply capacity, and Kress said Nvidia will continue working with TSMC as TSMC executives recently visited the company.
Source: IT DAILY · Kim Byeong-ju
Original: https://www.itdaily.kr/news/articleView.html?idxno=243008
References
This article was produced with the help of an automated content generation algorithm.
Source: IT DAILY
View originalThis article was summarized and organized by BizCrush based on the original article from IT DAILY. For exact quotations and full details, please refer to the original article.
