This morning Jensen Huang He brought his lucky T-shirt to wear, he chose from among the many leather jackets, all apparently the same, the one he had already worn in his golden days and even the glasses will have to have an impeccable light. If he has to play them all, today at 18:XNUMX Italian time, the ceo of Nvidia when he will take the stage at the GTC, the annual software developers conferenceThe event has always represented a crucial moment for the entire technology sector because it offers indications and reveals new models, systems, aggregations, the consistency of the market.
But this year's is much more: on the international table of technology, where at the head of the table reigned unchallenged The Magnificent Seven, including Nvidia, has been evident since the end of last January small chinese startup Deepseek and has proven to have created similar models, but at much lower prices, which were also joined by Chinese giant Alibaba who has developed an open source new product focused on improving reasoning skills.
Huang's words will be scrutinized: investors will have to Evaluating Nvidia's Future, with updates on the product roadmap and market prospects, but also on the entire US technology sector, characterized by extremely high capex expenses.
Le Nvidia shares listed on the Nasdaq closed yesterday at 119,53 dollars, with a decrease of 1,76%, but what is most impressive is the 11% decline since the beginning of the year. Huang will be called upon to defend his chip company's dominance from nearly 3 trillion dollars, as pressure mounts on its biggest customers to curb costs of artificial intelligence. The climate of uncertainty weighs on all the so-called "Magnificent Seven"Use, rename yourself"Evil Seven” by Goldman Sachs due to their heavy losses since the beginning of the year (over -10%), while the 7 Chinese tech heavyweights, the “7 Titans”, including Alibaba Group and Tencent, have gained 40% this year.
Nvidia shares have more than quadrupled in value in the last three years, as the company has fostered the rise of advanced artificial intelligence systems including ChatGPT or Claude. Much of this success is certainly due to the 10 years in which the Californian company has dedicated to building software tools to woo AI researchers and developers. But it was the Nvidia data center chips, which are sold at tens of thousands of dollars each, to represent the most of its $130,5 billion sales last year.
Anticipation for the new Rubin family
Huang is expected to reveal today new details on the most recent artificial intelligence chip of the company. Huang hinted that the new flagship offer it will be called Vera Rubin, with the product expected to enter mass production by the end of the year that will be part of a new chip family, reports Reuters, including a graphics processing unit, a central processing unit, and network chips – all designed to work in huge data centers that train AI systems. Analysts expect the chips to go into production this year and ship in high volumes starting next year.
Nvidia tries to introduce a new chip model every year, but so far it has encountered both internal and external obstacles. The company's current flagship chip, called Blackwell, is coming to market more slowly than expected, industry media reports, after a design flaw caused production problems. The AI industry at large has been grappling with delays over the past year as previous methods of feeding massive amounts of data into ever-larger data centers packed with Nvidia chips have begun to show diminishing returns.
Huang challenges DeepSeek
To DeepSeek's suggestions that its products have much less computing power and therefore fewer Nvidia chips, Huang responded that in reality new AI models that spend more time thinking about their answers will make Nvidia's chips even more important, because they are the fastest at generating “tokens,” the fundamental unit of AI programs. “When ChatGPT was born, the token generation speed had to be equal to the reading speed,” Huang told Reuters last month. “But the token generation rate now is the rate at which the AI can read itself, because it’s thinking about itself. And the AI can think about itself much faster than it and I can read, and because it has to generate so many future possibilities before it comes up with the right answer.”
Is Nvidia a Buy Now?
After the stellar earnings of the past few years and the collapse of a few weeks, the question is spreading among investors but is it time to buy Nvidia stock again? According to some analysts, Nvidia could be a buying opportunity. Kevin Cassidy Rosenblatt Securities expects GTC to be a positive catalyst for the stock, thanks to updates in networking and software, two of the company’s most profitable areas. Melius Research says Nvidia will need to demonstrate that its AI investments (including AI for drug discovery and AI agents for work) will continue to generate value over the long term. The custom GPU (ASIC) market is a threat, but Nvidia could remain at least a generation ahead of its competitors.
