Nvidia Unveils AI Advances, Stock Up 43% Yearly

Nvidia, a leading technology company, made significant announcements at this year's Computex Taipei, unveiling developments ranging from humanoid robotics to advanced NVLink technology. This technology allows enterprises to build semi-customized AI servers using Nvidia's infrastructure. These announcements come at a time when Nvidia's popularity is surging, following the U.S. decision to repeal the Biden administration's AI proliferation rules, which had restricted certain countries from purchasing the company's AI chips. Nvidia was also a key topic during Trump's visit to Saudi Arabia, where the company pledged to supply tens of thousands of AI processors to Humain, an AI startup backed by the Saudi sovereign wealth fund, over the next five years.

During the event, Nvidia introduced the Nvidia Isaac GR00T-Dreams, a tool designed to help developers generate vast amounts of training data. This data is used to teach robots to perform various tasks and adapt to new environments. Nvidia's CEO, Jensen Huang, has previously stated that physical AI represents the next trillion-dollar industry. To achieve this goal, the company is focusing on developing the necessary software to train and operate humanoid robots in factories, with the ultimate aim of bringing them into homes.

In addition to robotics, Nvidia showcased the new NVLink Fusion product, which allows customers to use Nvidia's Grace CPU and third-party AI chips in conjunction with Nvidia's various server infrastructure products to build customized servers. Customers also have the option to pair their own CPUs with Nvidia's AI chips. Nvidia stated in a press release, "Through NVLink Fusion, supercomputing customers can collaborate with Nvidia's partner ecosystem to integrate Nvidia's rack-level solutions, enabling seamless deployment in data center infrastructure." The idea is to provide infrastructure customers with more options when building data centers and server systems.

To further this initiative, Nvidia is developing the RTX Pro Blackwell servers. These servers, powered by Nvidia's Blackwell server version GPU, aim to drive the transition "from CPU-based systems to efficient GPU-accelerated infrastructure." Nvidia explains that these systems are designed to run "almost all enterprise workloads," including applications ranging from design and simulation software to running proxy AI programs.

Nvidia also launched the DGX Cloud Lepton, which allows customers to access cloud-based AI processing capabilities, enabling them to develop and deploy their own AI software. This is made possible through collaborations with partners such as CoreWeave, Foxconn, and SoftBank, who will host the global GPU cloud network.

These announcements come after a period of uncertainty for Nvidia, which has faced concerns over semiconductor tariffs, new export restrictions on chip shipments, and reports of Microsoft canceling some data center leases. Despite these challenges, Nvidia's stock has shown resilience, remaining flat year-to-date and experiencing a slight decline over the past six months. However, over the past 12 months, Nvidia's stock has risen by over 43%.