Nvidia GTC 2024: What to expect from the AI giant’s big conference

To say Nvidia (NVDA) is on a hot streak would be an understatement. Shares of the AI darling are up a staggering 267% over the last 12 months and 79% year to date. Tech companies across the world covet its graphics cards like rare jewels and CEO Jensen Huang is as in-demand as some heads of state.

And on Monday, he’ll kick off Nvidia’s annual GTC conference with a two-hour keynote at the SAP Center in San Jose, Calif., outlining what’s ahead for the company in the year ahead. In prior years, Nvidia has used the show to debut some of its biggest products.

In 2022, it announced its Hopper graphics architecture and H100 graphics processing unit, which is now the go-to card for companies training and deploying AI models. This year, Nvidia is widely expected to debut Hopper and the H100’s successors, kicking off what could be a new rush on the company’s products.

The new architecture, codenamed Blackwell, and GPU, dubbed the B100, are anticipated to offer far better performance when it comes to running models like OpenAI’s GPT-3. Nvidia isn’t exactly hiding its plans for the B100. During the Supercomputer 2023 conference, the company showed off a slide outlining the GPU’s potential capabilities compared to the H100 and H200 cards.

In addition to the Blackwell architecture and B100 card, Nvidia will also likely give attendees and viewers a look at the latest advancements in its CUDA software.

FILE - Nvidia Co-founder, President, and CEO Jensen Huang speaks at the Taiwan Semiconductor Manufacturing Company facility under construction in Phoenix, Tuesday, Dec. 6, 2022. Nvidia shares skyrocketed early Thursday after the chipmaker forecast a huge jump in revenue for the next quarter, notably pointing to chip demand for AI-related products and services.(AP Photo/Ross D. Franklin, File)
Nvidia CEO Jensen Huang will take the stage at GTC 2024 on March 18 in San Jose. (Ross D. Franklin/AP Photo, File) (ASSOCIATED PRESS)

The platform, which lets programmers take advantage of a GPU’s processing capabilities for AI and other applications, is an essential part of Nvidia’s overall enterprise strategy and helps lock in developers who build on Nvidia’s software, virtually ensuring they’ll stick with Nvidia’s products in the future.

Nvidia could also announce that it will begin producing its H20 AI chip for the Chinese market. The chip, according to Reuters, is meant to serve as an alternative to Nvidia’s more powerful H100 and H200 chips. The US prohibits Nvidia from selling those chips to Chinese customers for fear that the Chinese military and government will use them to develop advanced AI applications to rival the US’s.

Those are the big stories we’ll be watching from GTC, but we’ll also likely get plenty of other announcements and debuts from Nvidia and its partners.

Some 300 exhibitors are set to take part in the conference and speakers ranging from OpenAI’s COO Brad Lightcap and Meta’s vice president of AI research Joelle Pineau to Microsoft’s vice president of generative AI Sébastien Bubeck set to make appearances at the show.

Leave a Reply

Your email address will not be published. Required fields are marked *