Nvidia (NVDA) Founder and CEO Jensen Huang on Monday kicked off the company's annual GPU Technology Conference (GTC) with a keynote address, unveiling the Blackwell platform and engaging with industry leaders. We announced expanded partnerships and showcased technology capabilities.
Nvidia's event ran through Thursday and was eagerly anticipated by investors looking to learn about the chipmaker's artificial intelligence efforts and partnerships. NVIDIA stock has risen more than 80% since the beginning of the year on optimism about the company's AI-powered growth prospects.
Here are some key takeaways from Huang's presentation and the company's announcements on the first day of the event.
Introducing Blackwell, “The World’s Most Powerful Chip”
Nvidia has announced the latest version of its AI-powered technology, the Blackwell platform. The company said Blackwell GPUs are “the world's most powerful chips” and will power “a new era of computing.”
The new system is more capable and efficient than the previous Hopper.
“Blackwell's inference power is extraordinary, in fact, about 30 times greater than Hopper,” Huang said in his keynote.
The Blackwell architecture can run large language models for generative AI at approximately 25 times less cost and energy than Hopper, Nvidia said in a release.
Nvidia also announced the GB200 Grace Blackwell Superchip, a “massive superchip” that “connects two NVIDIA B200 Tensor Core GPUs to an NVIDIA Grace CPU.”
“The GB200 is also now available on NVIDIA DGX™ Cloud, an AI platform co-developed with leading cloud service providers, providing enterprise developers with the infrastructure and software they need to build and deploy advanced generative AI models,” the company said. “This will give you exclusive access to.”
Nvidia reports that a number of leading AI companies are already planning to use Blackwell systems ahead of its official launch.
Amazon (AMZN), Microsoft (MSFT), and Alphabet's Google (GOOGL) are among the partners planning to use Blackwell.
Nvidia said Blackwell products will be available through partners later this year.
Nvidia's Omniverse meets Apple's VisionPro
Nvidia has announced that it will bring Omniverse to Apple's (AAPL) recently launched VisionPro virtual reality headset.
Omniverse is Nvidia's “computing platform that enables individuals and teams to develop 3D workflows and applications.” Huang began his keynote with a demonstration of omniverse simulation.
Developers can stream Universal Scene Description (OpenUSD) industrial scenes from content creation applications to the NVIDIA Graphics Delivery Network (GDN), a global network of graphics-enabled data centers that can stream advanced 3D experiences to Apple Vision Pro. It’s easy to send,” NVIDIA said.
Next is AI humanoid robot
Nvidia announced Project GR00T, a basic model of a humanoid robot, which Huang said is the “next wave” of AI and robotics.
“Constructing a foundational model for a general humanoid robot is one of the most exciting problems to solve in AI today,” said Huang, adding that “leading robotics researchers around the world are The enabling technologies are coming together to take a giant leap forward.”
“Robots equipped with GR00T, which stands for Generalist Robot 00 Technology, will be designed to understand natural language and emulate movements by observing human behavior,” the company said.
Nvidia also announced Jetson Thor, a new computer for humanoid robots with GenAI capabilities as part of its AI-powered robotics initiative.
Blueprint for the “next generation” data center
Nvidia has presented a 3D blueprint for a “next generation” data center for the AI era.
The design was created using Omniverse to simulate a “modern data center” that is “highly complex, with multiple considerations such as performance, energy efficiency, and scalability,” the company said. Ta.
“In the future, as I mentioned earlier, data centers will be thought of as AI factories,'' Huang said, adding that the goal of an AI factory is to “generate intelligence.'' added.
Nvidia's latest products are expected to significantly enhance data centers. “By introducing GB200 Grace Blackwell processors to his Microsoft data centers around the world,” Microsoft CEO Satya Nadella said in a release. [the company is] build on [its] We have a long history of optimizing NVIDIA GPUs for our cloud, making the promise of AI a reality for organizations everywhere. ”
Nvidia and Oracle expand sovereign AI partnership
Nvidia announced it is expanding its collaboration with Oracle (ORCL) to support the global delivery of sovereign AI.
Sovereign AI “refers to the ability of nations to use their own infrastructure, data, workforce, and business networks to create artificial intelligence,” Nvidia explained.
Nvidia and Oracle's partnership will “combine Oracle's distributed cloud, AI infrastructure, and generative AI services with NVIDIA's accelerated computing and generative AI software to help governments and enterprises use AI factories.”
“As AI reshapes business, industry, and policy around the world, countries and organizations must strengthen their digital sovereignty to protect their most valuable data,” Oracle CEO Safra Catz said in a release. There is a need to do so.”
New 6G cloud platform AI
Nvidia announced a 6G research platform that provides “a comprehensive suite for researchers to advance AI in radio access network (RAN) technologies.”
“This platform enables organizations to accelerate the development of 6G technology that connects trillions of devices to cloud infrastructure, creating a hyper-intelligent world supported by self-driving cars, smart spaces, and a wide range of augmented reality and immersive educational experiences. “and collaborative robots,'' the company said in a release.
Nvidia said the platform powers the “next stage of wireless technology” and can meet the growing demand created by running AI on wireless devices.
Nvidia shares, which were up 0.7% during Monday's regular trading hours, fell 1.8% to $868.90 in extended trading.
