Tencent and Baidu, two of China’s largest technology companies, revealed how they’re keeping in the global artificial intelligence race even as the U.S. tightens some curbs on key semiconductors.
The business’ methods include stockpiling chips, making AI models more efficient and even using homegrown semiconductors.
Big names in the sector addressed the issue during their latest earnings conference calls.
Martin Lau, president of Tencent — the operator of China’s biggest messaging app WeChat — said his company has a “pretty strong stockpile” of chips that it has previously purchased. He was referring to graphics processing units (GPUs), a type of semiconductor that has become the gold standard for training huge AI models.
These models require powerful computing power supplied by GPUs to process high volumes of data.
But, Lau said, contrary to American companies’ belief that GPU clusters need to expand to create more advanced AI, Tencent is able to achieve good training results with a smaller group of such chips.
“That actually sort of helped us to look at our existing inventory of high-end chips and say, we should have enough high-end chips to continue our training of models for a few more generations going forward,” Lau said.
Regarding inferencing — the process of actually carrying out an AI task rather than just training — Lau said Tencent is using “software optimization” to improve efficiency, in order to deploy the same amount of GPUs to execute a particular function.
Lau added the company is also looking into using smaller models that don’t require such large computing power. Tencent also said it can make use of custom-designed chips and semiconductors currently available in China.
“I think there are a lot of ways [in] which we can fulfill the expanding and growing inference needs, and we just need to sort of keep exploring these venues and spend probably more time on the software side, rather than just brute force buying GPUs,” Lau said.
Baidu’s approach
Baidu, China’s biggest search company, touted what it calls its “full-stack” capabilities — the combination of its cloud computing infrastructure, AI models and the actual applications based on those models, such as its ERNIE chatbot.
“Even without access to the most advanced chips, our unique full stack AI capabilities enable us to build strong applications and deliver meaningful value,” Dou Shen, president of Baidu’s AI cloud business, said on the company’s earnings call this week.
Baidu also touted software optimization and the ability to bring down the cost of running its models, because it owns much of the technology in that stack. Baidu management also spoke about efficiencies that allow it to get more out of the GPUs it possesses.
“With foundation models driving up the need for a massive computing power, the abilities to build and manage large scale GPU clusters and to utilize GPUs effectively has become key competitive advantages,” Shen said.
The Baidu executive also touted the progress made by domestic Chinese technology firms in AI semiconductors, a move he said would help mitigate the impact of U.S. chip curbs.
“Domestically developed self-sufficient chips, along with [an] increasingly efficient home-grown software stack, will jointly form a strong foundation for long-term innovation in China’s AI ecosystem,” Shen said.
China domestic chip focus
China has been ramping up development of chips designed and manufactured on its home soil for the last few years. Most experts agree that Beijing remains overall behind the U.S. in the realm of GPUs and AI chips, but there have been some advances.
Gaurav Gupta, an analyst covering semiconductors at Gartner, said stockpiling is one way Chinese companies are dealing with export restrictions. Additionally, there has been some progress made in semiconductor technology in China, even if it remains behind the U.S., Gupta added.
“China has also been developing its own domestic semiconductor ecosystem, all the way from materials to equipment to chips and packaging. Different segments have made varying levels of progress, but China has been surprisingly extremely consistent and ambitious in this goal, and one must admit that they have achieved decent success,” Gupta told CNBC by email.
“This provides an avenue for them to procure AI chips, which perhaps can’t compete with those from the U.S chip leaders but continue to make progress.”
Many U.S. executives have urged Washington to scrap export restrictions in light of China’s progress. Nvidia CEO Jensen Huang called the curbs a “failure” this week, saying they are doing more damage to American businesses than to China.
Meta CEO Mark Zuckerberg appears at the Meta Connect event in Menlo Park, California, on Sept. 25, 2024.
David Paul Morris | Bloomberg | Getty Images
Meta’s AI assistant now has 1 billion monthly active users across the company’s family of apps, CEO Mark Zuckerberg said Wednesday at the company’s annual shareholder meeting.
The “focus for this year is deepening the experience and making Meta AI the leading personal AI with an emphasis on personalization, voice conversations and entertainment,” Zuckerberg said.
The artificial intelligent assistant’s 1 billion milestone comes after the company in April released a standalone app for the tool.
The plan is for Meta to keep growing the product before building a business around it, Zuckerberg said on Wednesday. As Meta AI improves overtime, Zuckerberg said “there will be opportunities to either insert paid recommendations” or offer “a subscription service so that people can pay to use more compute.”
In February, CNBC reported that Meta was planning to debut a standalone Meta AI app during the second quarter and test a paid-subscription service akin to rival chat apps like OpenAI’s ChatGPT.
“It may seem kind of funny that a billion monthly actives doesn’t seem like it’s at scale for us, but that’s where we’re at,” Zuckerberg told shareholders.
During the Meta shareholder meeting, investors voted on 14 different items related to the company’s business, nine of which were shareholder proposals covering topics such as child safety, greenhouse gas emissions and a proposed bitcoin treasury assessment.
Shareholder proposal 8, for example, was submitted by JLens, which is an investment advisor and affiliate of the Anti-Defamation League, and called for Meta to prepare an annual report detailing and addressing hate content, including antisemitism, on its services following January policy changes that relaxed content-moderation guidelines.
Early voting results on Wednesday showed the proposals that Meta’s board did not recommend were unlikely to pass, including one calling for the company to end its dual-class share structure, which gives Zuckerberg significant voting power. Meanwhile, the voting items that the board favored, including those pertaining to approving the company’s board of director nominees and an equity incentive plan, were likely to pass, based on the preliminary results.
Meta said final polling results will be released within four business days on the company’s website and the U.S. Securities and Exchange Commission.
Salesforce CEO Marc Benioff participates in an interview at the World Economic Forum in Davos, Switzerland, on Jan. 22, 2025.
Chris Ratcliffe | Bloomberg | Getty Images
Salesforce shares were volatile in extended trading on Wednesday after the sales and customer service software maker reported upbeat fiscal first-quarter results and guidance.
Here’s how the company performed relative to LSEG consensus:
Earnings per share: $2.58 adjusted vs. 2.54 expected
Revenue: $9.83 billion vs. $9.75 billion expected
Salesforce’s revenue grew 7.6% year over year in the quarter, which ended on April 30, according to a statement. Net income of $1.54 billion, or $1.59 per share, was basically flat compared with $1.53 billion, or $1.56 per share, a year ago.
President Donald Trump announced sweeping tariffs on goods imported into the U.S. in early April. Co-founder and CEO Marc Benioff sounded positive about the company’s results for the quarter anyway, pointing to its plan, announced on Tuesday, to buy data management company Informatica for $8 billion.
It would be Salesforce’s priciest acquisition since the $27.1 billion Slack deal in 2021. Slack marked the top end of the buyouts Salesforce had made under Benioff. Activist investors raised concerns about all the spending, in addition to slowing revenue growth.
Salesforce sprung into action, slashing 10% of its headcount. Benioff proclaimed that the board’s mergers and acquisitions committee had been disbanded. The company’s finance chief at the time said it would reach a margin expansion goal two years early. And Salesforce started paying dividends to shareholders.
Initial reception to the Informatica announcement was generally favorable. “Salesforce is paying a reasonable multiple for the asset, in our view, and the deal should be more easily digested by investors than some of the company’s large deals in the past (i.e. Slack),” Stifel analysts led by J. Parker Lane wrote in a note to clients. The investment bank has a buy rating on Salesforce shares.
During the fiscal first quarter, Salesforce introduced the AgentExchange marketplace for artificial intelligence agents.
Management sees $2.76 to $2.78 in adjusted earnings per share on $10.11 billion to $10.16 billion in revenue for the fiscal second quarter. Analysts polled by LSEG had expected $2.73 in adjusted earnings per share on $10.01 billion in revenue.
Salesforce bumped up its full-year forecast. It called for $11.27 to $11.33 in adjusted earnings per share and $41.0 billion to $41.3 billion in revenue, implying revenue growth between 8% and 9%. The LSEG consensus included net income of $11.16 per share and $40.82 billion in revenue. The guidance in February was $11.09 to $11.17 in adjusted earnings per share, with $40.5 billion to $40.9 billion in revenue.
As of Wednesday’s close, the stock had slipped about 18% so far in 2025, while the S&P index was unchanged.
Executives will discuss the results with analysts on a conference call starting at 5 p.m. ET.
This is breaking news. Please check back for updates.
HP reported second-quarter results that beat analysts’ estimates for revenue but missed on earnings and guidance, in part due to President Donald Trump’s sweeping tariffs. Shares sank 15% after the report.
Here’s how the company did versus analysts’ estimates compiled by LSEG:
Earnings per share: 71 cents adjusted vs. 80 cents expected
Revenue: $13.22 billion vs. $13.14 billion expected.
Revenue for the quarter increased 3.3% from $12.8 billion in the same period last year. HP reported net income of $406 million, or 42 cents per share, down from $607 million, or 61 cents per share, a year ago.
For its third quarter, HP said it expects to report adjusted earnings of 68 cents to 80 cents per share, missing the average analyst estimate of 90 cents, according to LSEG. Full-year adjusted earnings will be within the range of $3 to $3.30 per share, while analysts were expecting $3.49 per share.
HP said its outlook “reflects the added cost driven by the current U.S. tariffs,” as well as the associated mitigations.
“While results in the quarter were impacted by a dynamic regulatory environment, we responded quickly to accelerate the expansion of our manufacturing footprint and further reduce our cost structure,” HP CEO Enrique Lores said in a statement.
Lores told CNBC’s Steve Kovach that HP has increased production in Vietnam, Thailand, India, Mexico and the U.S. By the end of June, Lores said the company expects nearly all of its products sold in North America will be built outside of China.
“Through our actions, we expect to fully mitigate the increased trade-related costs by Q4,” Lores said in the interview.
HP will hold its quarterly call with investors at 5 p.m. ET.