Nvidia’s annual AI conference, GTC, is set to begin on Monday, with the event running through Friday and will be held in San Jose. The conference is expected to witness a gathering of tech enthusiasts eager for fresh developments in accelerated computing, AI, and robotics.
Aside from a wide variety of activities that will take place at the event, Nvidia’s Jensen Huang is expected to give his keynote address at the SAP Center by 10am Pacific Time. Many of the sessions will be available to viewers online, with the keynote streamed on Nvidia’s website and YouTube channel.
Nvidia’s GTC 2025 will reveal new GPUs and discuss quantum computing
According to Nvidia, the conference is expected to witness about 1,000 sessions, with 2,000 keynote speakers and about 400 exhibitors gracing the event. The event will witness discussions around sovereign AI, automotive innovation, and the latest in AI agents. Nvidia is also expected to reveal most of its AI GPUs during the event. Analysts have predicted that the company could provide news relating to new hardware for memory-intensive AI training.
Insiders also expect the spotlight on Blackwell, the firm’s next major chip family. Huang confirmed during the company’s latest earnings call that the Blackwell B300 series, code-named Blackwell Ultra, will be available in the second half of this year. These cards offer not just a boost in computing speed but also a jump in memory capacity, packing 288GB. This upgrade could be attractive to AI research involving data-hungry models.
However, the Blackwell lineup is only a part of the story, with Nvidia’s Rubin series, scheduled for release in 2026, is rumored to provide what Huang has called a “big, big, huge step up” in processing power. In addition, Nvidia has scheduled a “quantum day” to explore quantum computing progress and its implications for real-world use.
Meanwhile, the swift progress of Chinese AI lab DeepSeek, which developed efficient models that rival offerings from established AI firms, has caused some investors to wonder if demand for Nvidia’s highest-tier GPUs might reduce. However, Huang believes that DeepSeek’s emergence will broaden the overall AI market, raising the need for robust computing solutions. He also points to the rise of elaborate “reasoning” models, such as OpenAI’s o1, as a catalyst for more powerful GPU requirements.
Despite investor worries, Nvidia’s financial performance was relatively better. Nvidia announced record financial results in February. It reported $39.3 billion in revenue and predicted $43 billion for the following quarter. Rival chipmakers like AMD have started to grab some market share, but Nvidia still commands roughly 82% of
the GPU segment.