If you’ve checked your Amazon cart lately and done a double-take at the price of a simple 32GB kit, you aren’t hallucinating. As of early 2026, we are officially in the middle of a global memory crisis that analysts are calling “RAMageddon.”
Prices for DDR5 memory have spiked by over 100% to 300% in the last six months alone. That “budget” 32GB kit that was $80 in 2024? You’re likely seeing it listed for $350 or more today.
It isn’t just corporate greed—though the “Big Three” (Samsung, SK Hynix, and Micron) are certainly reporting record profits. It’s a structural shift in how silicon is being used.
- The AI Data Center Feast: Large Language Models (LLMs) and AI clusters require massive amounts of HBM (High Bandwidth Memory). Because HBM is far more profitable than the RAM in your laptop, manufacturers have pivoted nearly 70% of their production capacity to serve giants like Microsoft, Google, and OpenAI.
- Cannibalization of Wafers: HBM4 production is “wafer-hungry.” It uses up to 3.5x more silicon area than standard DDR5. Every chip sold to a data center is a chip that isn’t going into your gaming rig.
- The Death of DDR4: Manufacturers are aggressively EOL-ing (End of Life) DDR4 production to make room for DDR5 and HBM. This has caused a secondary price spike in older memory as supply evaporates for people trying to maintain “legacy” systems from 2023.