Top story for the morning: Nvidia beats forecasts, Amazon prepares up to $50 billion for OpenAI

Investors are used to Nvidia's results exceeding Wall Street's expectations / Photo: JHVEPhoto / Shutterstock.com
Nvidia reported record revenue of $68.1 billion last quarter thanks to the data center boom and forecasts up to $79.6 billion for the current period. The company finally received a U.S. license for limited shipments of its H200 artificial intelligence chips to China. And Amazon is discussing a major investment in developer ChatGPT. These and other topics - in the review of key events for the morning of February 26.
Nvidia beats earnings expectations
Nvidia reported a strong fourth quarter: revenues totaled $68.1 billion with EPS of $1.62, above analysts' forecasts, Yahoo Finance reports. The data center business ($62.3 billion), especially sales to hyperscalers, was the main contributor; the computing segment grew 58% year-over-year and networking solutions grew 263%. The first-quarter forecast ($76.4-79.6 billion) was also higher than expected and does not take into account possible revenues from China.
Despite this, the stock has barely risen: after a brief 3% rise in the postmarket, the growth has come to naught, the publication notes. Investors are questioning whether the AI boom will continue or whether the market is close to a peak, although Nvidia's stock has been outperforming AMD and Broadcom since the beginning of the year, second only to Intel.
OpenAI poached a top AI researcher from Meta
OpenAI has hired Ruomin Pan, one of Meta's key AI infrastructure specialists who previously joined the company from Apple, The Information reports. At Meta, he oversaw the infrastructure of Superintelligence Labs, which is developing the next generation of AI models. According to the publication, he was actively poached for several months and left Meta last week.
Pan's compensation package at Meta exceeded $200 million for several years, the publication points out. His move reflects the escalating "war for talent" in Silicon Valley, where tech giants are offering multimillion-dollar contracts in the race for AI leadership.
Amazon could invest up to $50 billion in OpenAI, subject to an IPO or a breakout in AGI
Amazon is discussing an investment of up to $50 billion in OpenAI: $15 billion could be provided immediately, and another $35 billion could be provided if the company reaches the AGI (Artificial General Intelligence - artificial general intelligence, i.e. a system capable of thinking and learning at the human level in a wide range of tasks) milestone or goes public, The Information reports. In parallel, SoftBank and Nvidia plan to invest $30 billion each within a year in the current round.
OpenAI is paving the way for a possible IPO with a valuation of up to $1 trillion, with big tech companies looking to solidify partnerships with the startup, which is investing heavily in data centers, to gain an edge in the AI race. In a full deal, Amazon could be the largest investor in the round.
Nvidia has received a U.S. license to ship the H200 to China
Nvidia said it has received a license from U.S. authorities for limited shipments of less advanced H200 chips to China - with mandatory inspection and a 25% duty, Bloomberg writes. However, the company has not yet received the revenue and does not know whether Beijing will allow the imports. Revenue from Chinese data centers is not included in the first-quarter outlook.
Sales of more advanced chips to China remain restricted for national security reasons, and the PRC AI chip market itself could reach $50 billion in the coming years, according to estimates by company head Jensen Huang. In parallel, Chinese authorities are supporting local manufacturers - Huawei, Cambricon, MetaX and Moore Threads - which, according to Nvidia CFO Colette Kress, could eventually change the global balance in the AI industry.
Alphabet integrates Intrinsic into Google to boost AI robotics
Alphabet has decided to incorporate the former "lunar" startup Intrinsic into Google's structure in an effort to simplify and accelerate the development of robotics, CNBC writes. The company is developing the Flowstate platform, a web service for creating robotic applications without thousands of lines of code. The technology will now utilize Gemini models, Google Cloud infrastructure and work closely with Google's DeepMind teams.
The move moves robotics software into the category of key corporate priorities and intensifies competition with Amazon and Tesla. Intrinsic is also already working with Foxconn and Nvidia to deploy AI robots on assembly lines in the US. According to Intrinsic CEO Wendy Tan White, the integration with Google will scale "physical AI" to a wider range of manufacturing companies.
What's in the markets
- Japan's broad Topix index was up 0.92 percent, while the Nikkei 225 was up 0.10 percent.
- Hong Kong's Hang Seng index was down 0.80%. CSI 300 index of mainland China - by 0.25%.
- In South Korea, the Kospi index added 2.52% and the Kosdaq added 4.07%.
- Australia's S&P/ASX 200 was up 0.51%.
- S&P 500 futures were down 0.22%, Nasdaq Composite futures were down 0.33% and Dow Jones Industrial Average exchange-traded contracts were down 0.24%.
This article was AI-translated and verified by a human editor
