Practical Strategies for Navigating the Memory Shortage

The semiconductor industry is facing an unprecedented memory shortage spanning high-bandwidth memory (HBM), DRAM, and NVMe flash, with no relief expected until 2027. DRAM prices have increased 171% year-over-year, and procurement lead times have stretched to 6-12 months.

Most organizations respond by trying to buy more storage. But the teams thriving through this shortage are the ones who figure out how to do more with what they have and tripling GPU productivity in the process.

Join experts from NAND Research and WEKA to understand what’s driving the shortage and how leading organizations are navigating it without delaying strategic AI initiatives.

In this webinar, you’ll learn:

• What’s really happening in the memory market from NAND Research insights on manufacturer capacity shifts, pricing trends, and timeline expectations.
• How to use what you already have by transforming underutilized NVMe resources in your GPU servers into memory-class performance, deploying in weeks [or days?] instead of waiting months for separate storage hardware you can’t procure
• Why software-over-silicon wins during shortages and how production deployments are achieving 3x greater GPU productivity, 1,000x GPU memory extension, and 20x time-to-first-token improvements using resources already in their deployment pipeline

Date: Tuesday, Feb 24
Time: 10am PT / 1pm ET
Location: Virtual – Join from Anywhere

Meet Our Speakers

Val Bercovici

Chief AI Officer, WEKA

Steve McDowell

Senior Analyst & Founder, NAND Research