Nvidia (NVDA, Financial), a global leader in computing technology, recently opened its "AI Summit" in Washington, D.C. The three-day event aims to showcase to policymakers and media how Nvidia's chip architecture and customizable AI services can benefit society. Unlike previous events, this summit is led by Bob Pette, Nvidia's VP and GM of Enterprise Platforms, focusing more on AI applications rather than new product launches.
Pette emphasized that the world is on the brink of widespread AI adoption across various industries. He projected that AI could create a $20 trillion impact across sectors utilizing the technology. The keynote highlighted the next phase of AI, with applications ranging from advanced AI agents to robotic factories, weather prediction, cancer treatment, and extraterrestrial exploration.
Thousands of pre-trained "AI agents," such as Nvidia NIM agents, are currently transforming productivity in industries worldwide. For instance, the U.S. National Cancer Institute uses Nvidia's medical AI services for medical imaging analysis and data extraction. Additionally, Nvidia's AI assists pharmaceutical companies in new drug molecule screening, significantly reducing development time.
Other AI agent applications include cybersecurity frameworks that monitor and respond to network threats in real-time and space exploration AI that detects faint radio signals from space. The "Digital Twin Earth" project, NVIDIA Earth 2, enables meteorologists to predict natural disasters more accurately and quickly.
Pette shared that Nvidia's CUDA library now features over 4,000 accelerated applications, facilitating breakthroughs across various sectors. According to Nvidia, the next AI wave will be "physical AI," involving robots interacting with the physical world while being controlled remotely.
Ultimately, these applications rely on Nvidia's computing power. As 2024 progresses into its final quarter, the highly anticipated GB200 AI servers are set for delivery. Microsoft Azure has disclosed that it has acquired GB200-equipped servers, becoming the first cloud service provider to implement the Blackwell system. OpenAI has also received initial shipments of Nvidia's DGX B200 engineering machines.
Nvidia previously projected shipments of billions of dollars' worth of new AI servers in the fourth quarter, driven by intense market demand. Following this week's event, Nvidia plans to host AI summits in Mumbai, India, and Tokyo, Japan, with both countries having government-backed large-scale computing infrastructure investment plans.
