NVIDIA at CoreWeave Fully Connected 2026
September 29–October 1 | San Francisco
Explore the NVIDIA AI Factory
NVIDIA and CoreWeave are redefining what’s possible in AI infrastructure—combining NVIDIA’s accelerated computing platform with CoreWeave’s purpose-built cloud to deliver the AI factory at scale. At Fully Connected, go deeper on the full stack: from the infrastructure powering today’s most ambitious models to training, inference, agentic, and physical AI. Hear from NVIDIA leaders, explore live demos, and see how the partnership is accelerating the next wave of AI breakthroughs.
Keynote Day 1 at Fully Connected: Built for Adoption
Watch the Keynote Day 1 at Fully Connected
Watch live at Fully Connected on September 30.
Ian Buck joins as a special guest for the opening Keynote at Fully Connected to share his perspective on the AI factory infrastructure powering today’s most demanding agentic AI workloads. His conversation will explore how NVIDIA and CoreWeave are partnering to deliver the performance that AI breakthroughs demand.
Keynote Speaker Ian Buck
VP of Hyperscale and High-Performance Computing, NVIDIA
NVIDIA Session Schedule at CoreWeave Fully Connected
Wednesday, September 30
2–2:45 p.m. PT
A Fireside Chat on Physical AI, Featuring Caterpillar
Learn how Caterpillar is using AI to solve real-world challenges and drive innovation. This session will share practical experiences, lessons learned, and best practices for building and scaling AI.
Brandon Hootman, VP of Physical AI, Caterpillar
Deepu Talla, Vice President, Robotics and Edge, NVIDIA
Wednesday, September 30
3–3:45 p.m. PT
Customizing Open Models for Agentic Workloads
Open models are becoming the foundation for agentic AI, but success depends on choosing, adapting, and optimizing the right model for each workflow. This session explores open models, including NVIDIA Nemotron™, and how developers can improve long-running agentic workflows over time with reinforcement learning, fine-tuning, and inference optimization.
Avinash Kaur, Product Lead, Gen AI Software Co-Design, NVIDIA
Wednesday, September 30
4–4:45 p.m. PT
Building Reliable AI Infrastructure at Scale With Extreme Co-Design
Reliable AI infrastructure at scale starts long before the first system arrives. In this session, learn how NVIDIA and CoreWeave are applying extreme co-design across compute, networking, software, and operations to bring up next-generation AI factories built for the NVIDIA Vera Rubin platform. We’ll explore how Vera CPUs, Rubin GPUs, NVLink™, ConnectX™, BlueField™, and Spectrum-X™ Ethernet work together to deliver predictable performance, resiliency, and efficiency for model builders and AI-native companies serving production AI applications.
Brian Sparks, Senior Director, Networking Product Marketing, NVIDIA
Kirthi K. Devleker, Senior Product Marketing Manager, NVIDIA
Thursday, October 1
2–2:45 p.m. PT
AI Tokenomics Explained
The economics of AI inference are reshaping how organizations plan, build, and scale their AI investments. This session unpacks the critical levers from silicon to software that determine AI token output and directly impact the bottom line. Through real-world use cases, gain a practical framework for sizing your AI infrastructure needs at scale. See how NVIDIA’s extreme co-design approach maximizes tokens-per-watt and minimizes cost-per-token and cost-per-task, enabling organizations to profitably scale AI revenue.
Shruti Koparkar, Senior Manager, Product Marketing, NVIDIA
Thursday, October 1
4–4:45 p.m. PT
NVIDIA Vera CPU: The Foundation for Agentic AI
Agentic AI demands high-performance CPUs. NVIDIA Vera is the CPU for agents, helping them call tools, execute code, and power reinforcement learning loops. CPU performance determines how quickly AI factories can turn intelligence into action. This session explores Vera’s agentic sandbox performance and its role in accelerating orchestration, data processing, and other CPU-intensive workloads that keep GPUs fed and agents responsive.
Hannah Coutand, Director of Product Marketing, Vera CPU, NVIDIA
Featured Speakers
Join Us
At Fully Connected, the full AI factory stack built on NVIDIA and CoreWeave comes to life, from infrastructure to agentic and physical AI.
September 29–October 1, 2026
NVIDIA Programs and Technical Training
NVIDIA Training
Our expert-led self-paced courses, instructor-led workshops, and technical certification provide learners with the knowledge and hands-on experience necessary to unlock the full potential of NVIDIA solutions.
NVIDIA Startup Program
The NVIDIA Inception program gives startups access to the latest developer resources, preferred pricing on NVIDIA software and hardware, and unique exposure to the venture capital community. The program is free and available to tech startups of all stages.
NVIDIA Developer Program
The NVIDIA Developer Program provides developers with comprehensive resources, tools, and a community of experts that help them build accelerated applications and solutions. Join and get a complimentary course from the Deep Learning Institute.
FAQ
NVIDIA at CoreWeave Fully Connected takes place from September 29 to October 1, 2026, in San Francisco, California. The event will showcase the full AI factory stack, from core infrastructure to agentic and physical AI.
Learn more about the event and speakers on the CoreWeave Fully Connected Event Page. Register to attend by visiting the Registration Portal.
The opening Keynote, “Keynote Day 1 at Fully Connected: Built for Adoption,” will feature special guest Ian Buck, VP of Hyperscale and High-Performance Computing at NVIDIA. He will discuss the AI factory infrastructure powering today’s most demanding AI workloads on September 30.
Learn more about the event and speakers on the CoreWeave Fully Connected Event Page.
The partnership combines NVIDIA’s accelerated computing platform with CoreWeave’s purpose-built cloud to deliver scalable AI factory infrastructure for model training, inference, agentic AI, and physical AI.
Explore details on the CoreWeave Fully Connected Event Page.
NVIDIA and CoreWeave apply extreme co-design across compute, networking, software, and operations to build next-generation AI factories powered by the NVIDIA Vera Rubin platform, incorporating Vera CPUs, Rubin GPUs, NVLink, ConnectX, BlueField, and Spectrum-X Ethernet.
View the full agenda on the CoreWeave Fully Connected Event Page.
You can subscribe to NVIDIA news directly on the event landing page or follow NVIDIA across social media platforms including Facebook, Twitter, LinkedIn, Instagram, and YouTube.
Stay connected via the CoreWeave Fully Connected Event Page.