– NVIDIA has announced its next-generation Blackwell GPU architecture for accelerated computing and real-time generative AI
– Blackwell promises up to 25 times lower cost and energy consumption compared to its predecessor
– The platform features a powerful chip, second-generation Transformer Engine, NVLink interconnect, and support from major cloud providers and partners
NVIDIA has introduced its next-generation Blackwell GPU architecture, aimed at advancing accelerated computing and enabling organizations to create and run real-time generative AI on trillion-parameter large language models. Named after mathematician David Harold Blackwell, this platform offers significant cost and energy savings compared to its predecessor, the Hopper architecture.
The Blackwell architecture includes innovative technologies such as the powerful GB200 Grace Blackwell Superchip and a fifth-generation NVLink interconnect for enhanced communication between GPUs. This new GPU architecture has garnered support from major cloud providers and partners like Amazon Web Services, Google Cloud, and Microsoft Azure, among others.
Industry leaders such as Jensen Huang of NVIDIA and Sundar Pichai of Alphabet and Google are optimistic about the potential of Blackwell to drive breakthroughs in various sectors including engineering, scientific computing, and generative AI. Meta CEO Mark Zuckerberg highlighted the importance of AI in powering future technologies and expressed excitement about leveraging Blackwell for their own AI projects.
With its high performance and efficiency, Blackwell aims to make real-time trillion-parameter AI a viable option for enterprises. The Blackwell architecture is set to revolutionize the AI landscape and has garnered interest from various industry players eager to explore its capabilities. Interested individuals can learn more about AI and big data at the AI & Big Data Expo events in Amsterdam, California, and London.