NVIDIA Unveils Blackwell Platform to Power the Generative AI Era
NVIDIA CEO Jensen Huang introduces the new Blackwell computing platform at GTC 2024 to handle trillion-parameter AI models. The announcement also includes NIM microservices and Omniverse Cloud APIs to accelerate enterprise AI adoption.
NVIDIA founder and CEO Jensen Huang kicks off the largest GTC conference to date by unveiling the Blackwell computing platform, a new processor designed specifically for the generative AI era. Addressing a massive crowd of over 12,000 in-person attendees and tens of thousands online, Huang declares that accelerated computing reaches a tipping point as traditional general-purpose computing runs out of steam.
The newly introduced Blackwell platform delivers a massive upgrade to global AI infrastructure by enabling real-time generative AI on trillion-parameter large language models. To help developers deploy custom AI models across hundreds of millions of GPUs, Huang also announces NIM, which stands for NVIDIA inference microservices, serving as a new method for packaging and delivering AI software.
In addition to processing power and software deployment, Huang introduces Omniverse Cloud APIs to bring advanced simulation capabilities into the physical world. These major announcements feature powerful demonstrations and partnerships with some of the world's largest enterprises, showcasing NVIDIA's plan to revolutionize industries ranging from software and robotics to medical technology.