Historical memory prices 1960-2026(dam.stanford.edu)
322 points by vga1 17 hours ago | 118 comments
tl;dr: An interactive dataset tracking historical memory prices ($/GB) from 1960 to 2026 across DRAM, NAND flash, and HBM, extending John C. McCallum's classic dataset with monthly updates from Keepa (Amazon retail) and quarterly HBM estimates from Epoch AI. It breaks out DRAM by generation (SDRAM through DDR5), HBM by generation (HBM2e through projected HBM4), and includes a modeled accelerator cost breakdown for Nvidia, AMD, Google TPU, and AWS Trainium. Raw CSV data is downloadable; caveats note that "cheapest retail" often reflects EOL clearance rather than leading-edge pricing.
HN Discussion:
  • ~Pre-1990 per-GB pricing is unrealistic and not inflation-adjusting distorts the historical picture
  • The data refutes memory manufacturers' claims that RAM/storage are no longer commodities
  • The DRAM data is misleading because recent points reflect EOL DDR3 rather than current generations
  • Curiosity about what drives the visible cyclical price patterns in the data
  • AI and crypto demand are driving price volatility and hurting consumers