Breaking Down the Memory Wall in
AI Infrastructure
AI is booming, but infrastructure limitations are creating a critical bottleneck: GPUs can process data incredibly fast, but can’t hold enough of it in memory (aka “the memory wall”). When AI deployments hit this constraint, efficiency and ROI plummet.
How will this impact your AI strategies in the short and long term? Join experts from WEKA and S&P Global/451 Research to discover how next-gen architectures break through the memory wall to power scalable, cost-efficient AI.
- Understand the challenge: Get key insights on infrastructure bottlenecks facing enterprises scaling AI initiatives
- Reimagine memory architecture: How a new “memory tier” built on NVMe storage is enabling petabyte-scale token persistence.
- Flip the economics of inference: Learn how sustained cache hit rates can deliver up to 90% lower costs and 10–100x greater GPU efficiency.
Date: Thursday, Nov 13, 2025
Time: 10AM PT | 1PM ET | 6PM GMT
Location: Virtual – Join from Anywhere!
10AM PT | 1PM ET | 6PM GMT
AI is booming, but infrastructure limitations are creating a critical bottleneck: GPUs can process data incredibly fast, but can’t hold enough of it in memory (aka “the memory wall”). When AI deployments hit this constraint, efficiency and ROI plummet.
How will this impact your AI strategies in the short and long term? Join experts from WEKA and S&P Global/451 Research to discover how next-gen architectures break through the memory wall to power scalable, cost-efficient AI.
- Understand the challenge: Get key insights on infrastructure bottlenecks facing enterprises scaling AI initiatives
- Reimagine memory architecture: How a new “memory tier” built on NVMe storage is enabling petabyte-scale token persistence.
- Flip the economics of inference: Learn how sustained cache hit rates can deliver up to 90% lower costs and 10–100x greater GPU efficiency.
Date: Thursday, Nov 13, 2025
Time: 10AM PT | 1PM ET | 6PM GMT
Location: Virtual – Join from Anywhere!
Speakers
-
Val Bercovici
AI Factories, WEKA
-
Greg Macatee
Senior Analyst, Lead AI Infrastructure, S&P Global, 451 Research