About Rambus Design Summit

The Rise of Agentic AI and Memory-Centric Computing

Speaker: Steven Woo

Agentic AI is ushering in a new era of computing where intelligent systems can reason, plan, and act with minimal human intervention. These workloads are driving unprecedented demands on memory capacity, bandwidth,latency, and data movement efficiency across the data center, edge, and client platforms. In this keynote, we explore how the rise of AI agents is reshaping system architectures and accelerating the shift toward memory-centric computing. Attendees will gain insight into the technology trends, infrastructure requirements, and innovations needed to support the next generation of autonomous AI systems.

Main Memory Solutions for AI Servers and Agentic Workloads

Speaker: Zaman Mollah

As AI workloads continue to grow in complexity, main memory has become a critical enabler of overall system performance. Agentic AI applications further increase demands for larger context windows, faster retrieval of information, and the ability to support multiple concurrent AI agents. This session examines the evolution of server memory architectures and the role of advanced memory solutions in delivering higher bandwidth, larger capacities, improved reliability, and greater power efficiency. Learn how innovations in memory subsystems are enabling the next generation of AI servers and data center platforms.​

AI PCs and Edge Agents: Memory Solutions for Local Intelligence

Speaker: Carlos Weissenberg

The emergence of AI PCs and intelligent edge devices is bringing advanced AI capabilities closer to users and data sources. Local AI processing requires memory subsystems that balance performance, power efficiency, and form factor constraints. This session explores the memory technologies enabling AI at the edge, including the role of DDR5 and LPDDR5 memory architectures in supporting increasingly capable on-device AI models and autonomous agents.
Attendees will gain insight into the challenges and opportunities of delivering local intelligence at scale.

How Agentic AI Raises the Stakes

Speakers: Steven Woo, John Eble, Lou Ternullo, and Tim Messegee

Agentic AI promises greater automation, productivity, and business value, but it also raises new challenges for system architects, infrastructure providers, and technology leaders. As AI agents operate with greater autonomy, infrastructure must support higher levels of performance, reliability, power efficiency and scalability. In this roundtable discussion, Rambus technology experts will examine how Agentic AI is changing requirements across memory and interconnect technologies. Join the conversation to explore emerging opportunities, critical design considerations, and the implications of autonomous AI for future computing platforms.

Memory IP for AI Accelerators: Solutions from the Data Center to Intelligent Edge

Speaker: Zachi Friedman

AI accelerators are becoming the engines behind modern AI workloads, requiring highly optimized memory interfaces to feed increasingly powerful compute resources. This session explores the latest memory IP solutions that enable high-bandwidth, low-latency data access across a wide range of AI applications, from hyperscale data centers to intelligent edge devices. Learn how advanced memory interfaces, subsystem architectures, and design techniques
can help accelerator developers maximize performance, reduce power consumption, and accelerate time-to-market for next-generation AI platforms.​ ​

Scale Up and Scale Out with PCIe 7

Speaker: Phil Swart​ As AI clusters grow beyond the boundaries of individual servers, high-performance interconnects have become essential for scaling compute resources efficiently. PCIe 7 represents a significant leap in bandwidth and connectivity, enabling next-generation AI systems to move data faster and support larger, more complex workloads. In this session, we examine how PCIe 7 technologies support both scale-up and scale-out architectures, improve accelerator connectivity, and help remove system bottlenecks. Attendees will learn how next-generation PCIe switching and interconnect solutions are shaping the future of AI infrastructure.​

Security Solutions for Protecting AI Hardware and Assets

Speaker: Vincent van der Leest​

As AI systems become more valuable and widely deployed, securing hardware, models, data, and infrastructure has become a mission-critical requirement. From data center deployments to edge AI devices, organizations must establish trust throughout the hardware lifecycle while protecting against increasingly sophisticated attacks. This session explores the latest security technologies, including hardware root of trust solutions, quantum-safe cryptographic protections, and standards-based security frameworks. Learn how comprehensive hardware security architectures help protect AI assets, ensure system integrity, and build trust in next-generation AI platforms.​