A look at the main points of NVIDIA Corporation's (NVDA.US) GTC speech by Huang Renxun: heavyweight collaborations including 6G, NVQLink integrating quantum computing and supercomputing, bursting the AI bubble, Blackwell's American-made.

On Tuesday, the 28th of September, NVIDIA Corporation CEO Huang Renxun delivered the keynote speech at the second GTC conference of the year in Washington D.C., covering technological breakthroughs in the areas of 6G, AI, quantum computing, and Siasun Robot & Automation. Huang Renxun emphasized in his speech that with Moore's Law becoming obsolete, accelerated computing and GPU technology have become the core driving force behind technological advancement. In the aspect of combining AI with 6G technology, NVIDIA Corporation announced a strategic partnership with Nokia Oyj Sponsored ADR, investing $1 billion in the subscription of Nokia Oyj Sponsored ADR shares to jointly promote an AI-native 6G network platform. In the field of supercomputing, NVIDIA Corporation introduced the NVQLink technology, which integrates AI supercomputing with quantum computing, connecting quantum processors with GPU supercomputers and receiving support from 17 quantum computing companies. NVIDIA Corporation also announced a collaboration with the U.S. Energy Corp. to build the largest AI supercomputer in the department. In the realm of AI factories, NVIDIA Corporation will be launching the Bluefield-4 processor to support AI factory operations. Furthermore, NVIDIA Corporation added fuel to the trend of Robotaxi services for autonomous vehicles, announcing partnerships with the ride-sharing pioneer Uber and the Chrysler parent company Stellantis. Uber plans to deploy 100,000 vehicles utilizing NVIDIA Corporation technology for Robotaxi services starting in 2027. NVIDIA Corporation also formed partnerships with the AI star Palantir and the pharmaceutical giant Lilly, integrating their GPU computing capabilities with enterprise data platforms and pharmaceutical R&D to drive AI from concept to practical application. These collaborations, targeting enterprise operations intelligence and drug development, signify the accelerated commercialization of AI technology in complex industry scenarios. Huang Renxun stated in his speech, "AI is the most powerful technology of our time, and science is its greatest frontier." The collaborations announced on Tuesday mark NVIDIA Corporation's strategic transformation from a chip manufacturer to a full-stack AI infrastructure provider. Huang Renxun refuted claims of an AI bubble, stating, "I don't believe we are in an AI bubble. We are using all of these different AI models - we are using a lot of services, and we are willing to pay for it." His central argument is that AI models are now powerful enough, and customers are willing to pay for them, which in turn proves that the expensive investment in computing infrastructure is justified. Chip shipments surge, capacity expansion accelerates Huang Renxun revealed that NVIDIA Corporation's fastest AI chip Blackwell GPU is now in full production in Arizona. This means that the previously Taiwan-manufactured Blackwell chip can now be manufactured in the United States. Huang Renxun disclosed staggering data on NVIDIA Corporation's chip shipments, expecting to ship 20 million Blackwell chips. In comparison, the previous-generation product, the Hopper architecture chip, only shipped 4 million units throughout its lifecycle. Huang Renxun also mentioned that 6 million Blackwell GPUs have been shipped in the past four quarters, with demand remaining strong. NVIDIA Corporation expects that Blackwell and the upcoming Rubin chip will collectively bring in $500 billion in GPU revenue over five quarters. Earlier this month, NVIDIA Corporation and Taiwan Semiconductor Manufacturing Co. Ltd. Sponsored ADR announced that the first batch of Blackwell wafers had been produced at a factory in Phoenix, Arizona. NVIDIA Corporation stated in a video that systems based on Blackwell are now also being assembled in the United States. NVIDIA Corporation teams up with Nokia Oyj Sponsored ADR for 6G network deployment NVIDIA Corporation, in partnership with Nokia Oyj Sponsored ADR, will launch the Aerial RAN Computer (ARC) to catalyze the transformation of 6G networks. NVIDIA Corporation and Nokia Oyj Sponsored ADR will pioneer AI platforms for 6G communication technology. How does 6G integrate with AI? In addition to AI learning and enhancing 6G spectral efficiency, we will also see AI-enhanced wireless access network (RAN) products, namely "AI on RAN". This means that, under the current internet status, much of the data is processed on Amazon.com, Inc.'s AWS cloud service platform, but NVIDIA Corporation aims to build a cloud computing platform on top of the 6G connection. This demonstrates the potential of ultra-fast AI, which can power technologies such as autonomous vehicles. On Tuesday, NVIDIA Corporation and Nokia Oyj Sponsored ADR announced a strategic partnership to incorporate NVIDIA Corporation-driven commercial AI-RAN products into Nokia Oyj Sponsored ADR's RAN product portfolio, enabling communication service providers to launch AI-native 5G-Advanced and 6G networks. NVIDIA Corporation will launch the Aerial RAN Computer Pro platform for 6G networks, and Nokia Oyj Sponsored ADR will expand its RAN product portfolio on this basis, introducing new AI-RAN products. NVIDIA Corporation will make a $1 billion equity investment in Nokia Oyj Sponsored ADR at a subscription price of $6.01 per share. Analytical firm Omdia predicts that the AI-RAN market size is expected to exceed $200 billion by 2030. The collaboration between NVIDIA Corporation and Nokia Oyj Sponsored ADR will provide distributed edge AI inference capabilities, opening up new high-growth areas for telecommunications operators. T-Mobile U.S. will collaborate with Nokia Oyj Sponsored ADR and NVIDIA Corporation to drive testing and development of AI-RAN technology, integrating the technology into its 6G development process. Trials are expected to begin in 2026, focusing on verifying performance and efficiency improvements. This technology will support AI-native devices such as autonomous vehicles, drones, augmented reality glasses, and virtual reality. NVQLink connects quantum computing with GPU systems Currently, various quantum computing technologies exhibit powerful performance but are sensitive to environmental noise, limiting their application range. This is where GPU-based supercomputers come into play, relieving the burden on quantum processors. On Tuesday, Huang Renxun mentioned that NVIDIA Corporation has built the open-source system architecture NVQLink based on its open-source quantum development platform CUDA-Q. Huang Renxun stated that, in addition to new technologies, quantum computing also requires support from traditional processors, and NVIDIA Corporation aims to help achieve this goal. "We now realize that directly connecting quantum computers to GPU supercomputers is crucial. This is the future of computing quantization." NVQLink is a new high-speed interconnection technology that connects quantum processors with GPUs and CPUs. It is not intended to replace quantum computers but to accelerate quantum computing in conjunction with them. Huang Renxun said that NVQLink technology will aid in error correction and calibrating which AI algorithms should be used on GPUs and quantum processors. He revealed that 17 quantum computing companies have committed to supporting NVQLink. "The industry's support is incredible. Quantum computing will not replace traditional systems; they will work in harmony." "It (NVQLink) can not only correct today's quantum bits, but also future ones. We will scale the quantum computers from the hundreds of quantum bits currently to tens of thousands of quantum bits, and even hundreds of thousands of quantum bits in the future." NVIDIA Corporation stated that the NVQLink technology has gained support from 17 quantum processor manufacturers and 5 controller manufacturers, including companies such as Alice & Bob, Atom Computing, IonQ, IQM Quantum Computers, Quantinuum, Rigetti, among others. Nine national laboratories led by the U.S. Energy Department will use NVQLink to advance quantum computing breakthroughs, including labs such as Bruker Corporation's Haverhill National Laboratory, Fermilab, and Los Alamos National Laboratory (LANL), among others. NVIDIA Corporation mentioned that developers can access NVQLink through the CUDA-Q software platform to create and test applications that seamlessly call CPUs, GPUs, and quantum processors. NVIDIA Corporation collaborates with Oracle Corporation to build the largest AI supercomputer for the U.S. Energy Department Huang Renxun announced that NVIDIA Corporation will partner with the U.S. Energy Department to build seven new supercomputers, to be deployed at the Argonne National Laboratory (ANL) and Los Alamos National Laboratory (LANL) under the department. NVIDIA Corporation has announced a collaboration with Oracle Corporation to build the department's largest AI supercomputer, the Solstice system, which will include a record-breaking 100,000 NVIDIA Corporation Blackwell GPUs. Another system named Equinox will contain 10,000 Blackwell GPUs and is expected to be operational in the first half of 2026. Both systems will be interconnected through the NVIDIA Corporation network, providing a total of 2200 exaflops of AI performance. These supercomputers will enable scientists and researchers to develop and train new cutting-edge models and AI inference models using NVIDIA Corporation's Megatron-Core library and expand them using the TensorRT inference software stack. Energy Secretary Chris Wright stated, "Maintaining America's leadership in high-performance computing requires us to build a bridge to the next computing era: accelerating quantum supercomputing. Deep collaboration between our national laboratories, startups, and industry partners such as NVIDIA Corporation is critical to this mission." Argonne National Laboratory Director Paul K. Kearns stated that these systems will seamlessly connect with advanced experimental facilities such as the Advanced Photon Source under the Energy Department, allowing scientists to address the nation's most pressing challenges through scientific discovery. BlueField-4 drives AI factory infrastructure upgrades Huang Renxun believes that agent-based AI is no longer just a tool but an assistant to all of our work. AI brings "countless opportunities." NVIDIA Corporation's plan is to build factories dedicated to AI, filled with chips. NVIDIA Corporation announced on Tuesday the launch of the Bluefield-4 processor that supports AI factory operating systems. NVIDIA Corporation's BlueField-4 data processing unit supports an 800Gb/s throughput, providing breakthrough acceleration for gigabit-level AI infrastructure. This platform combines NVIDIA Corporation's Grace CPU and ConnectX-9Network-1 Technologies, Inc., with computing power six times that of its predecessor BlueField-3, capable of supporting three times the scale of AI factories as BlueField-3. BlueField-4 is designed for a new class of AI storage platforms, laying the foundation for efficient data processing and breakthrough performance for AI data pipelines. The platform supports multi-tenant networks, fast data access, AI runtime security, and cloud elasticity, natively supporting NVIDIA Corporation's DOCA microservices. NVIDIA Corporation stated that several industry leaders are planning to adopt BlueField-4 technology. This includes companies in the server and storage sectors such as Cisco Systems, Inc., DDN, Dell Technologies, Inc. Class C, HPE, IBM, Lenovo, Supermicro, VAST Data, and WEKA. In the field of network security, companies such as Armis, Check Point, Cisco Systems, Inc., F5, Forescout, Palo Alto Networks, and Trend Micro are also planning to incorporate BlueField-4. Additionally, cloud and AI service providers such as Akamai, CoreWeave, Crusoe, Lambda, Oracle Corporation, Together.ai, and xAI are building solutions based on NVIDIA Corporation's DOCA microservices, speeding up multi-tenant networks, enhancing data movement speed, and bolstering the security of AI factories and supercomputing clouds. NVIDIA Corporation's BlueField-4 is set to be launched as an early version of the Vera Rubin platform in 2026. NVIDIA Corporation partners with CrowdStrike for AI cybersecurity development Huang Renxun announced that NVIDIA Corporation will collaborate with the cybersecurity company CrowdStrike on AI cybersecurity models. NVIDIA Corporation announced a strategic partnership with CrowdStrike to provide NVIDIA Corporation AI computing services on the CrowdStrike Falcon XDR platform. This partnership combines the Falcon platform's data with NVIDIA Corporation's GPU-optimized AI pipeline and software, including the new NVIDIA Corporation NIM microservices, enabling customers to create customized security generative AI models. According to CrowdStrike's 2024 Global Threat Report, the average time to breach has decreased to 62 minutes, with the fastest recorded attack just over 2 minutes. With modern attacks becoming faster and more complex, organizations require AI-driven security technology to gain the necessary speed and automation capabilities. Huang Renxun stated, "Cybersecurity is fundamentally a data problem - the more data enterprises can process, the more events they can detect and respond to. By combining NVIDIA Corporation's accelerated computing and generative AI with CrowdStrike cybersecurity, enterprises are provided unprecedented threat visibility." CrowdStrike will leverage NVIDIA Corporation's accelerated computing, Morpheus, and NIM microservices to introduce LLM-driven applications into enterprises. Combined with the Falcon platform's unique contextual data, customers will be able to address new use cases in specific areas, including improving threat hunting by processing PB-level logs, detecting supply chain attacks, identifying abnormal user behavior, and actively defending against emerging vulnerabilities. NVIDIA Corporation's new autonomous driving development platform assists Uber in deploying Robotaxi fleet Huang Renxun mentioned that NVIDIA Corporation's end-to-end autonomous driving platform DRIVE Hyperion is ready to launch cars for Robotaxi services. Global automakers such as Stellantis, Lucid, and Mercedes-Benz will utilize NVIDIA Corporation's new DRIVE AGX Hyperion 10 architecture to accelerate the development of autonomous driving technology. NVIDIA Corporation announced a partnership with Uber, utilizing the next-generation NVIDIA Corporation DRIVE AGX Hyperion 10 autonomous driving development platform and DRIVE AV software to expand the world's largest L4-level mobility network. NVIDIA Corporation will support Uber in gradually increasing its global fleet of autonomous vehicles to 100,000 starting from 2027. DRIVE AGX Hyperion 10 is a reference-grade production computer and sensor architecture that enables any vehicle to reach L4 level readiness. This platform allows automakers to build cars, trucks, and vans equipped with validated hardware and sensors that can host any compatible autonomous driving software. Huang Renxun stated, "Robotaxi marks the beginning of a global transportation transformation - making traffic safer, cleaner, and more efficient. We have created a framework with Uber to deploy autonomous driving fleets on a massive scale." Uber CEO Dara Khosrowshahi stated, "NVIDIA Corporation is a pillar of the AI era and is now leveraging this innovation to unleash L4 autonomous driving capability on a massive scale." Stellantis is developing the AV-Ready platform, specially optimized to support L4-level capabilities and meet the requirements for autonomous taxi services. These platforms will integrate NVIDIA Corporation's full-stack AI technology, further enhancing connectivity with Uber's global mobility ecosystem. Uber stated that Stellantis will be among the first manufacturers to provide Robotaxi vehicles, delivering at least 5,000 NVIDIA Corporation-driven Robotaxi cars for Uber's operations in the U.S. and internationally. Uber will be responsible for end-to-end fleet operations, including remote assistance, charging, cleaning, maintenance, and customer support for vehicles. Stellantis stated it will collaborate with Foxconn on hardware and system integration, with manufacturing planned to start in 2028. Operations will initially begin in the United States in partnership with Uber. Stellantis expects pilot projects and testing to gradually expand over the coming years. Lucid is advancing L4-level autonomous driving capabilities for its next-generation passenger vehicles, using the full-stack NVIDIA Corporation AV software on the DRIVE Hyperion platform to deliver the first batch of L4-level autonomous driving cars to customers. Mercedes-Benz is testing future collaborations based on its proprietary MB.OS operating system and DRIVE AGX Hyperion, with the new S-Class models offering an exceptional L4-level driving and riding experience. NVIDIA Corporation and Uber will continue to support and accelerate global partners developing software stacks on the NVIDIA Corporation DRIVE L4 platform. Partners include Avride, May Mobility, Momenta, Nuro, Pony.ai, Wayve, and WeRide, for the autonomous vehicle transportation sector, Aurora, Volvo Autonomous Solutions, and Waabi are developing L4-level autonomous driving trucks powered by NVIDIA Corporation's DRIVE platform. NVIDIA Corporation collaborates with Palantir to develop AI technology stack Lowe's leads in applying supply chain optimization solutions NVIDIA Corporation and Palantir's collaboration focuses on integrating NVIDIA Corporation's GPU-accelerated computing, open-source models, and data processing capabilities into Palantir's AI platform (AIP) Ontology system. Ontology creates a digital twin for enterprises by organizing complex data and logic into interconnected virtual objects, links, and actions, providing the foundation for AI-driven business process automation. Huang Renxun stated, "Palantir and NVIDIA Corporation share a common vision: to operationalize AI, transforming enterprise data into decision intelligence. By combining Palantir's powerful AI-driven platform with NVIDIA Corporation's CUDA-X accelerated computing and Nemotron open-source AI models, we are building the next-generation engine to empower specialized AI applications and agents running some of the world's most complex industrial and operational pipelines." At a technical level, customers can use Ontology to process data using NVIDIA Corporation's CUDA-X data science libraries, coupled with NVIDIA Corporation's accelerated computing, to drive real-time AI decisions for complex business workflows. NVIDIA Corporation's AI Enterprise platform (including the cuOpt decision optimization software) will support enterprises in dynamic supply chain management. NVIDIA Corporation's Nemotron inference model and NeMo Retriever open-source model will help enterprises quickly build AI agents that leverage information from Ontology. Palantir co-founder and CEO Alex Karp stated, "Palantir focuses on deploying AI that can deliver immediate asymmetric value for customers. We are honored to collaborate with NVIDIA Corporation to integrate our AI-driven decision intelligence system with the world's most advanced AI infrastructure." Retailer Lowe's is one of the first enterprises to adopt the integrated technology stack from Palantir and NVIDIA Corporation, creating a digital twin of its global supply chain network to achieve dynamic and continuous AI optimization. The technology aims to enhance supply chain agility while achieving cost savings and customer satisfaction. Lowe's Chief Digital and Information Officer Seemantini Godbole remarked, "Modern supply chains are highly complex dynamic systems, and AI is crucial to helping Lowe's adapt and optimize quickly under changing conditions. Even minor changes in demand can have a cascading effect throughout a global network. By combining Palantir technology with NVIDIA Corporation AI, Lowe's is reimagining retail logistics to better serve our customers every day." NVIDIA Corporation and Palantir also plan to introduce the NVIDIA Corporation Blackwell architecture into the Palantir AIP, accelerating end-to-end AI pipelines from data processing and analysis to model development, fine-tuning, and production AI. Enterprises will be able to run AIP on NVIDIA Corporation's AI factory to achieve optimized acceleration. Palantir AIP will also receive support from NVIDIA Corporation's recently launched government AI factory reference design. Lilly collaborates with NVIDIA Corporation to build the most powerful supercomputer in the pharmaceutical industry, powered by over a thousand Blackwell Ultra GPUs. These chips will be connected through a unified high-speed network. The supercomputer will provide power for an AI factory, a dedicated computing infrastructure developed at Lilly to extensively train and deploy AI models for drug discovery and development. Lilly's Chief Information & Digital Officer, Diogo Rau, stated that it typically takes around 10 years from the first human trial of a drug to market launch. The company expects to complete the construction of the supercomputer and AI factory in December and launch it in January next year. However, the returns on these new tools may not be significant for Lilly and other pharmaceutical companies until late 2030. Rau said, "What we're talking about discovering with this kind of compute [are] the benefits that we can really start observing in 2030." Multiple AI models will be available on Lilly's TuneLab, an AI and machine learning platform that allows biotech companies to access AI models trained by Lilly based on its proprietary research over several years. This