Nvidia stormed San Jose this year, with a record of 25,000 participants who flock to the San Jose Convention Center and surrounding the buildings in the city center. Many workshops, conferences and panels were so excited that people had to rely against the walls or sit on the floor – and suffer from the anger of the organizers shouting orders to align them properly.
Nvidia is currently at the AI top of the world, with record finances, beneficiary margins from top to bottom and no serious competitor. But the coming months also have an unprecedented risk for the company, because it faces American prices, Deepseek and changing priorities of the main AI customers.
At the GTC 2025, the CEO of Nvidia, Jensen Huang, tried to project confidence, to reveal powerful new chips, personal “superordinators” and, of course, really cute robots. It was an exhaustive sale argument – one of the investors in shock from Nvidia’s nose actions.
“The more you buy, the more you save,” said Huang at some point in a speech on Tuesday. “It’s even better than that. Now, the more you buy, the more you do it. ”
Inferences boom
More than anything, NVIDIA, of this year’s GTC, sought to provide participants – and the rest of the world looking – that demand for his tokens will not slow down anytime.
During his speech, Huang said that almost the “whole world was mistaken” on the traditional scaling of the AI falling from the vogue. The Chinese AI laboratory Deepseek, which published a very effective “reasoning” model earlier this year called R1, has sparked fears among investors that Nvidia monster tokens can no longer be necessary to form a competitive AI.
But Huang has repeatedly insisted on the fact that the reasoning models swallowed to be able to stimulate more demand for the company’s chips, no less. This is why at GTC, Huang showed the next Line of the GPUs of Vera Rubin of Nvidia, saying that they will make the inference (that is to say the AI models) at around double the rate of the current Best Blackwell chip from Nvidia.
The threat to NVIDIA’s activities, Huang, spent less time attacking was parasites such as deceptions, groq and other low -cost equipment and cloud suppliers. Almost every hyperscaler develops a personalized chip for inference, if not training, too. AWS A Graviton and Inferentia (What would be aggressively), Google has TPU and Microsoft A Cobalt 100.
Along the same vein, technology giants are currently extremely dependent on Nvidia fleas, including Openai and Meta, seek to reduce these links via internal material efforts. If they – and the other aforementioned rivals – succeed, it will almost certainly weaken Nvidia’s workforce on the AI fleas market.
This may be why Nvidia’s action has dropped around 4% after Huang’s opening. Investors may have kept the hope of “one last thing”-or perhaps an accelerated launch window. In the end, they are not obtained neither.
Pricing
NVIDIA also sought to lie down prices at the GTC 2025.
The United States has not imposed any price on Taiwan (where Nvidia obtains most of its tokens), and Huang said that the prices would not do “significant damage” in the short term. He ceased to promise that Nvidia would however be protected from long -term economic impacts – regardless of the form they finally take.
Nvidia clearly received the “America First” message from the Trump administration, with Huang Promising in GTC To spend hundreds of billions of dollars for manufacturing in the United States, while it would help the company diversify its supply chains, it is also a massive cost for Nvidia, whose evaluation of several dollars of several dollars depends on healthy beneficiary margins.
New business
While trying to sow and develop companies other than its range of basic chips, NVIDIA de GTC has drawn attention to its new investments in Quantum, an industry that the company has historically neglected. During the first quantum day of the GTC, Huang apologized to the CEOs of large quantum companies for provoking a minor stock accident in January 2025 after suggesting that technology would not be very useful for the next 15 to 30 years.
Tuesday, nvidia announcement That he would open a new center in Boston, NVAQC, to advance quantum IT in collaboration with the “main” hardware and software markers. The center will of course be equipped with NVIDIA chips, which, according to the company, will allow researchers to simulate the quantum systems and the models necessary for quantum error correction.
In the more immediate future, Nvidia sees what he calls “AI personal supervisors” as a new potential income manufacturer.
At GTC, the company launched DGX Spark (previously called Project Digits) and DGX Station, which are both designed to allow prototyper users, refine and run AI models in a range of sizes at the edge. Neither is exactly cheap – they sell thousands of dollars – but Huang has boldly proclaimed that they represent the future of the personal PC.
“This is the AI age computer,” said Huang during his speech. “This is what computers should look like, and this is what computers will perform in the future.”
We will soon see if customers agree.
