NVIDIA · Video summary
Advancing Infrastructure for the Era of Agentic AI | Ian Buck at AI Infra Summit 2026
NVIDIA's Ian Buck discusses the evolution of AI infrastructure, highlighting the shift towards agentic AI and the full-stack Vera Rubin platform designed to meet its demanding computational needs.
Watch on YouTube ↗Key takeaways
- 💡The core of value creation in computing lies in the transformation of data through calculation, not just storage or communication.01:12
- 🌐NVIDIA's CUDA platform has expanded immensely over 26 years, now supporting over 3 million AI models and 8,000 major applications across diverse sectors from hyperscale to edge computing.01:35
- 🚀Agentic AI workloads are significantly more demanding than traditional chat AI, requiring 100 times more computational power due to dynamic input sizes, larger KV cache sizes, and an explosion in the number of turns.06:10
- 📈The NVIDIA Vera Rubin platform, through full-stack co-design, delivers up to 30x better AI factory throughput on agentic workloads, with some cases seeing up to 60x improvement.11:32
- ⚡NVIDIA's Vera CPU, featuring a custom Olympus core, is designed for agentic computing, offering 2x performance and 40% lower memory latency compared to other CPUs, crucial for fast tool calling and efficient agent responses.15:58
- 🔗NVLink Fusion enables seamless integration of custom accelerators (XPUs) into NVIDIA's AI infrastructure, allowing chip designers to leverage the entire NVLink ecosystem for scalable and high-performance AI factories.22:29
Chapters
- 0:33The Enormous Universe of Computing
- 5:25NVIDIA Vera Rubin: A Fungible, Full-Stack AI Factory Platform
- 10:10Agentic AI: 100x More Demanding Than Chat
- 15:11NVIDIA Vera Rubin Extreme Co-design for Agents
- 18:37Full-Stack Co-design Delivers Better AI Factory Throughput
- 22:30Groq 3 LPX Delivers Long Context with Extreme Token Speed
- 25:20NVIDIA Vera CPU for Agents
- 33:15Vera Delivers 1.5x More Agentic CPU Sessions for Perplexity
- 33:55NVIDIA Vera—The Fastest CPU for ClickHouse
- 35:25Vera Is Here and Shipping to Customers
- 35:46NVIDIA NVLink Enables Leading Performance for AI Factories
- 37:09NVLink Fusion Connects XPUs to NVIDIA AI Infrastructure
- 38:25d-Matrix Partners with NVIDIA on NVLink Fusion
- 38:53Amazon’s Annapurna Labs Partners with NVIDIA on NVHBM
- 40:29AI Factories Have Unused Power Capacity
- 43:29NVIDIA DSX MaxLPS Enables 40% More Factory Revenue
- 45:31AI Clouds Deploy NVIDIA DSX, Lambda Adopts MaxLPS
- 46:44NVIDIA DSX AI Factory Platform
The digest
Keep reading in The Digest
The full 1-minute article, and every new episode from NVIDIA and the creators you follow.
Read the full digest — freeMore from NVIDIA
The Infrastructure of Intelligence: NVIDIA and OpenAI on a Decade of Building Frontier AI
The AI Infra Summit brought together industry leaders to discuss the challenges and innovations in building AI infrastructure at scale. The conversation delved into the evolving definition of scale, the complexities of designing compute token factories, and the pivotal role of partnerships in navigating this unprecedented growth.
How Mistral Is Building Frontier AI for the Enterprise | NVIDIA AI Podcast Ep. 301
In this episode of the NVIDIA AI Podcast, Tim Lacroix, co-founder and CTO of Mistral AI, discusses the company's commitment to open source models, their collaboration with NVIDIA, and the development of their new platform, Mistral Forge, aimed at enhancing enterprise AI solutions. Lacroix emphasizes the importance of customization in AI to meet specific industry needs while fostering global innovation.
NVIDIA GTC 2026: Healthcare Special Address
In this NVIDIA GTC 2026 address, NVIDIA discusses the transformative impact of AI in healthcare, highlighting advancements in digital labs, agentic AI, and robotics that enhance disease understanding, drug discovery, and patient care. The presentation emphasizes collaboration and the integration of AI technologies to create a more efficient and accessible healthcare system.
Automotive Special Address: Advancing Level 4 Autonomy, the Path to Scalable, Safe AVs and Robotaxis
NVIDIA's Drive AV platform is evolving with physical AI and a three-computer architecture (training, simulation, inference) to accelerate Level 4 autonomy through hardware, software, and open-sourced models.
NVIDIA GTC 2026 Open Models Panel Highlights with Jensen Huang
AI is evolving beyond single models to complex systems and agents, with open models playing a crucial role in driving innovation, customization, and trust, especially for mission-critical applications.
NVIDIA GTC Telecom Special Address: The AI Grid—Intelligently Connecting AI Infrastructure
The AI Grid is a new telecom infrastructure connecting the physical world to AI, offering a generational opportunity for telcos by merging AI and telecom stacks for unprecedented scale and new monetization models.