NVIDIA has launched the Blackwell platform, representing a significant advancement in computing capabilities. This new platform is designed to empower organizations to develop and utilize real-time generative AI models with trillion-parameter scale, while reducing costs and energy consumption by up to 25 times compared to its predecessor.
The Blackwell platform introduces six innovative technologies for accelerated computing, which promise to drive breakthroughs across various industries such as data processing, engineering simulation, electronic design automation, computer-aided drug design, and quantum computing. NVIDIA’s CEO, Jensen Huang, expressed excitement about the potential of generative AI, stating that Blackwell will serve as the engine to power this new industrial revolution. He emphasized NVIDIA’s commitment to partnering with dynamic companies worldwide to harness the potential of AI across all sectors.
Major industry players, including Amazon Web Services, Google, Meta, Microsoft, OpenAI, Oracle, Tesla, and others, are expected to adopt the Blackwell platform. Sundar Pichai of Google highlighted the company’s longstanding partnership with NVIDIA and expressed enthusiasm for integrating the Blackwell GPU into Google Cloud services to accelerate future discoveries.
Similarly, Andy Jassy of Amazon emphasized the collaborative efforts between AWS and NVIDIA to develop innovative solutions such as Project Ceiba, which combines NVIDIA’s Grace Blackwell Superchips with AWS Nitro System for AI research and development. Michael Dell of Dell Technologies underscored the importance of generative AI in driving innovation and affirmed the company’s commitment to delivering next-generation accelerated products and services to customers.
Other industry leaders, including Demis Hassabis of Google DeepMind, Mark Zuckerberg of Meta, Satya Nadella of Microsoft, and Sam Altman of OpenAI, expressed their enthusiasm for the Blackwell platform and its potential to advance AI capabilities.
Named after mathematician David Harold Blackwell, the Blackwell architecture succeeds the NVIDIA Hopper architecture and introduces several groundbreaking technologies. These include the world’s most powerful chip, a second-generation transformer engine, fifth-generation NVLink, RAS engine, secure AI capabilities, and a dedicated decompression engine.
The NVIDIA GB200 Grace Blackwell Superchip, featuring two NVIDIA B200 Tensor Core GPUs interconnected by NVLink, offers superior AI performance and energy efficiency. The platform also includes the GB200 NVL72, a rack-scale system combining 36 Grace Blackwell Superchips with NVIDIA BlueField-3 data processing units, providing up to 30 times better performance for LLM inference workloads.
Blackwell-based products will be available from various partners, including major cloud service providers like AWS, Google Cloud, Microsoft Azure, and Oracle Cloud Infrastructure. Additionally, leading hardware manufacturers such as Cisco, Dell, Hewlett Packard Enterprise, Lenovo, and others will offer servers based on Blackwell technology.
Overall, the Blackwell platform represents a significant milestone in the evolution of computing, offering unprecedented capabilities for accelerating AI research and applications across industries. With widespread adoption anticipated across the technology ecosystem, the Blackwell platform is poised to drive transformative advancements in AI and computing.
