There is something for every technical level and interest area at GTC, with over 900 stimulating presentations, 300+ displays, 20+ technical seminars covering generative AI and more, and loads of unique networking activities. NVIDIA CEO Jensen Huang’s speech is highly anticipated.
Highlights From NVIDIA’s GTC24 Keynote
- Unveiling Blackwell, the ‘World’s Most Powerful Chip
- AI Humanoid Robots Are Next- “Project GR00T”
- Nvidia launches NIM to make it smoother to deploy AI models into production
- Nvidia’s Omniverse Meets Apple’s VisionPro
- Blueprint for ‘Next Gen’ Data Centers
- Nvidia and Oracle Expand Sovereign AI Partnership
- New 6G Cloud Platform AI
- Nvidia CEO wants enterprise to think ‘AI factory,’ not data center
Read: How to Incorporate Generative AI Into Your Marketing Technology Stack
Unveiling Blackwell, the ‘World’s Most Powerful Chip
As part of its ongoing commitment to developing cutting-edge AI technologies, Nvidia has unveiled the Blackwell platform. As the “world’s most powerful chip,” the Blackwell GPU will enable “a new era of computing,” according to the business. Following in Hopper’s footsteps, the new system is both more powerful and more efficient. Compared to Hopper, the Blackwell architecture uses around 25 times less energy and costs 25 times less to run generative AI large language models, according to Nvidia’s press statement. According to the business, “GB200 will also be available on NVIDIA DGX™ Cloud, an AI platform co-engineered with leading cloud service providers that gives enterprise developers dedicated access to the infrastructure and software needed to build and deploy advanced generative AI models.” Several prominent AI businesses are already planning to utilize the Blackwell system, according to Nvidia, even before it officially launches. Blackwell is going to be used by several partners, including Amazon (AMZN), Microsoft (MSFT), and Google (GOOGL), which is owned by Alphabet. Nvidia has announced that its partners will be allowed to sell Blackwell products later this year.
AI Humanoid Robots Are Next-“Project GR00T”
According to Huang, the “next wave” of artificial intelligence and robotics is Nvidia’s “Project GR00T,” a model for humanoid robots. According to Huang, the enabling technologies are coming together for leading roboticists around the world to take giant leaps towards artificial general robotics.” He went on to say that “building foundation models for general humanoid robots is one of the most exciting problems to solve in AI today. According to the business, robots powered by GR00T, which stands for Generalist Robot 00 Technology, will be designed to understand natural language and emulate movements by observing human actions. As part of its AI-powered robotics strategy, Nvidia also introduced “Jetson Thor,” a new microprocessor for humanoid robots that incorporates GenAI capabilities.
Nvidia launches NIM to make it smoother to deploy AI models into production
The new software platform, Nvidia NIM, was unveiled today at Nvidia’s GTC conference. Its purpose is to facilitate the integration of both custom and pre-trained AI models into production settings. NIM aggregates Nvidia’s software work on inferencing and optimizing models into a microservice that can be quickly and simply deployed by combining a model with an optimized inferencing engine, packaging it into a container, and releasing it into the wild. Nvidia claims that, even if the company had artificial intelligence (AI) experts on staff, it would take weeks or months for developers to release containers of this kind. Companies looking to accelerate their AI roadmap may take advantage of Nvidia’s curated microservices as the backbone of their AI-ready containers, which are built on top of Nvidia’s hardware. NVIDIA, A121, Adept, Cohere, Getty Images, Shutterstock, Google, Hugging Face, Meta, Microsoft, Mistral AI, and Stability AI are among the models that NIM presently supports. Currently, Nvidia is collaborating with Amazon Web Services (AWS), Google Cloud Platform (GCP), and Microsoft Azure AI to bring these NIM microservices to market. Deepset, LangChain, and LlamaIndex are just a few of the frameworks that will incorporate them.
Nvidia’s Omniverse Meets Apple’s VisionPro
The announcement came from Nvidia, who would be adding Omniverse support to Apple’s brand-new VisionPro VR headset. One of Nvidia’s products, the Omniverse, is a “computing platform that enables individuals and teams to develop 3D workflows and applications.”A demonstration of Omniverse simulations was the first part of Huang’s keynote. Software engineers may easily send their Universal Scene Description (OpenUSD) industrial scenes from their content creation applications to the NVIDIA Graphics Delivery Network (GDN), a global network of graphics-ready data centers that can stream advanced 3D experiences to Apple Vision Pro.
Read: 10 AI In Energy Management Trends To Look Out For In 2024
Nvidia and Oracle Expand Sovereign AI Partnership
To aid with the worldwide delivery of sovereign AI, Nvidia, and Oracle (ORCL) have announced that their partnership will be growing. The agreement between Nvidia and Oracle aimed to “deliver sovereign AI solutions to customers around the world” According to “Oracle’s distributed cloud, AI infrastructure, and generative AI services, combined with NVIDIA’s accelerated computing and generative AI software, are enabling governments and enterprises to deploy AI factories.”
New: 10 AI ML In Personal Healthcare Trends To Look Out For In 2024
New 6G Cloud Platform AI
A 6G research platform, touted by Nvidia, provides “researchers with a comprehensive suite to advance AI for radio access network (RAN) technology. The platform allows organizations to accelerate the development of 6G technologies that will connect trillions of devices with the cloud infrastructures, laying the foundation for a hyper-intelligent world supported by autonomous vehicles, smart spaces, and a wide range of extended reality and immersive education experiences and collaborative robots, according to a release from the company. The platform, according to Nvidia, will enable “the next phase of wireless technology” and may be able to satisfy the increasing demand for artificial intelligence (AI) on mobile devices.
Blueprint for ‘Next Gen’ Data Centers
In the age of artificial intelligence, Nvidia unveiled a 3D model of their “next-gen” data centers. Omniverse was used to model “modern data centers,” which are “incredibly complex, involving multiple considerations like performance, energy efficiency, and scalability,” according to the business, that developed the concept. It is believed that data centers will be significantly enhanced by Nvidia’s newest products. “By bringing the GB200 Grace Blackwell processor” to Microsoft datacenters globally, as stated in a release by Microsoft CEO Satya Nadella, “we are making the promise of AI real for organizations everywhere.” This announcement builds on Microsoft’s long history of optimizing NVIDIA GPUs for our cloud.
Read: Intel’s Automotive Innovation At CES 2024
Nvidia CEO wants enterprise to think ‘AI factory,’ not data center
Nvidia CEO Jensen Huang used the term “AI factory” multiple times in Monday’s keynote speech at the GTC 2024 event. Following his presentation, Huang spoke with TechCrunch about the last Industrial Revolution and how water was the primary input for factories. “Energy was the result.” The idea of data centers, which are little more than money pits, was like this, he said; they take raw materials and transform them into something else with value. “These [server] rooms are undergoing a new Industrial Revolution. I refer to them as AI factories,” Huang remarked. Information and power are the primary inputs. It produces data tokens as an output. Distributed globally, the token is undetectable. It is quite valuable.
[To share your insights with us as part of editorial or sponsored content, please write to sghosh@martechseries.com]
Comments are closed.