Disaggregated Memory System Architectures for AI and large database workloads
Thursday, June 25, 2026 2:40 PM to 3:00 PM · 20 min. (Europe/Berlin)
Hall H, Booth L01 - Ground Floor
HPC Solutions Forum
AI Applications powered by HPC TechnologiesIndustrial Use Cases of HPC, ML and QCMemory Technologies and HierarchiesSovereignty in AIStorage Technologies and Architectures
Information
Unlock the next level of performance for AI/ML and HPC workloads with disaggregated memory architecture.
Discover how CXL-based memory systems and fabric-attached memory (FAM) are transforming data center design—enabling dynamic, shared memory pools across servers without changing existing applications.
Built on Crete 2.0 (Micron + PNNL/DOE) and evolving into the production-ready Abaco 3.0 (Q4’2026), this approach delivers:
Massive scaling — Tens to hundreds of TB of near-memory, on demand
Breakthrough performance — Eliminating storage bottlenecks
AI-ready architecture — Optimized for LLM inference, vector & graph databases
Join this session to see how composable memory infrastructure is reshaping data centers—and powering the next generation of AI.
Discover how CXL-based memory systems and fabric-attached memory (FAM) are transforming data center design—enabling dynamic, shared memory pools across servers without changing existing applications.
Built on Crete 2.0 (Micron + PNNL/DOE) and evolving into the production-ready Abaco 3.0 (Q4’2026), this approach delivers:
Massive scaling — Tens to hundreds of TB of near-memory, on demand
Breakthrough performance — Eliminating storage bottlenecks
AI-ready architecture — Optimized for LLM inference, vector & graph databases
Join this session to see how composable memory infrastructure is reshaping data centers—and powering the next generation of AI.
HPC Solutions Forum Questions
What is the best way to keep advancing HPC in an AI-driven world?Why or when does it matter to have data or computation on-premises or in the cloud? Should it be all one or the other? How does your solution help?Discuss your solution in terms of benefits for specific use cases, rather than general horizontal terms like HPC, AI, performance, or scalability.
Format
on-site
