NVIDIA has unveiled its latest milestone in technological advancement with the launch of the NVIDIA Blackwell platform.
This innovative platform empowers organizations worldwide to harness the power of real-time generative AI on trillion-parameter large language models (LLMs), all while consuming up to 25 times less cost and energy than its predecessor.
The Blackwell GPU architecture incorporates six transformative technologies for accelerated computing, poised to revolutionize data processing, engineering simulation, electronic design automation, computer-aided drug design, quantum computing, and generative AI. NVIDIA sees these emerging industry opportunities as pivotal areas for the application of its cutting-edge technology.
Jensen Huang, founder, and CEO of NVIDIA, expressed the company’s longstanding commitment to accelerated computing and the significant role it plays in transformative breakthroughs like deep learning and AI. He emphasized generative AI as the defining technology of the present era, with Blackwell serving as the engine to power this new industrial revolution.
Jensen Huang highlighted NVIDIA’s collaboration with dynamic companies worldwide to realize the promise of AI across various industries.
Numerous organizations, including Amazon Web Services, Dell Technologies, Google, Meta, Microsoft, OpenAI, Oracle, Tesla, and xAI, are anticipated to adopt the Blackwell platform.
Sundar Pichai, CEO of Alphabet and Google, praised the deep partnership between Google and NVIDIA, expressing excitement about bringing the capabilities of the Blackwell GPU to Google’s Cloud customers and internal teams, including Google DeepMind.
Andy Jassy, President, and CEO of Amazon, emphasized the longstanding collaboration between AWS and NVIDIA, spanning over 13 years, and highlighted the extensive support for NVIDIA GPUs on AWS, including the co-development of Project Ceiba.
Michael Dell, founder, and CEO of Dell Technologies, emphasized the critical role of generative AI in creating smarter and more efficient systems, underscoring the partnership between Dell Technologies and NVIDIA in shaping the future of technology.
Demis Hassabis, Co-founder, and CEO of Google DeepMind, highlighted the transformative potential of AI in solving critical scientific problems and the crucial compute capabilities of Blackwell in facilitating new scientific discoveries.
Mark Zuckerberg, founder, and CEO of Meta, emphasized the importance of AI in powering various Meta products and expressed eagerness to leverage NVIDIA’s Blackwell for training open-source Llama models and advancing Meta AI and consumer products.
Satya Nadella, Executive Chairman, and CEO of Microsoft, underscored Microsoft’s commitment to offering advanced infrastructure for AI workloads and announced plans to integrate the GB200 Grace Blackwell processor into Microsoft’s data centers globally.
Sam Altman, CEO of OpenAI, lauded the significant performance leaps offered by Blackwell and expressed excitement about collaborating with NVIDIA to enhance AI compute.
Larry Ellison, Chairman, and CTO of Oracle, emphasized the collaboration between Oracle and NVIDIA in enabling breakthroughs in AI, machine learning, and data analytics, highlighting the need for powerful engines like Blackwell for accelerated computing and generative AI.
Elon Musk, CEO of Tesla and xAI, praised NVIDIA hardware as the best option for AI applications.
Named after David Harold Blackwell, the first Black scholar inducted into the National Academy of Sciences, the Blackwell architecture succeeds the NVIDIA Hopper architecture.
Blackwell Innovations to Fuel Accelerated Computing and Generative AI The Blackwell platform introduces six revolutionary technologies, enabling AI training and real-time LLM inference for models scaling up to 10 trillion parameters. These technologies include:
World’s Most Powerful Chip: Packed with 208 billion transistors, Blackwell-architecture GPUs are manufactured using a custom-built 4NP TSMC process.
Second-Generation Transformer Engine: Supports double the compute and model sizes with new 4-bit floating point AI inference capabilities.
Fifth-Generation NVLink: Delivers groundbreaking 1.8TB/s bidirectional throughput per GPU for seamless high-speed communication among up to 576 GPUs.
RAS Engine: Enhances reliability, availability, and serviceability, maximizing system uptime and reducing operating costs.
Secure AI: Protects AI models and customer data without compromising performance.
Decompression Engine: Accelerates database queries for enhanced data analytics and data science performance.
A Massive Superchip The NVIDIA GB200 Grace Blackwell Superchip connects two NVIDIA B200 Tensor Core GPUs to the NVIDIA Grace CPU over a 900GB/s ultra-low-power NVLink chip-to-chip interconnect.
The platform acts as a single GPU with 1.4 exaflops of AI performance and 30TB of fast memory, serving as a building block for the newest DGX SuperPOD.
Global Network of Blackwell Partners Blackwell-based products will be available from various partners, including cloud service providers, hardware manufacturers, and software makers.
NVIDIA offers robust software support through NVIDIA AI Enterprise, the end-to-end operating system for production-grade AI.
Anticipating market challenges in 2024, analysts project a potential drop in Nvidia’s market share as competitors introduce new products. However, Nvidia’s foray into software solutions, including microservices for AI efficiency and 3-D modeling tools, demonstrates its commitment to diversifying its offerings beyond hardware.
