We’re launching Lyria 3.5 in Google Flow Music, with advances across musicality, lyrics, vocals, and creative control
deepmind.google·6d ago
AI Engineering
Hacker News·Jun 28·5 min read
Historical memory prices 1960-2026
TL;DR
Historically, there has been a lack of accessible data on memory prices, which are crucial for hardware budgeting. David Shim created an interactive dataset that tracks the prices of DRAM, HBM, and NAND flash memory from 1960 to 2026.
✦ Why It Matters
Engineers can leverage this dataset to optimize hardware budgeting and forecast future memory costs.
Key Takeaways
- Utilize the dataset for budgeting hardware projects.
- Analyze price trends for better purchasing decisions.
- Download raw data for custom analysis.
- Stay updated on quarterly HBM price changes.
Full Summary
Memory prices have fluctuated significantly over the decades, impacting hardware development and cost management. David Shim developed an interactive dataset that compiles historical and current prices for DRAM (Dynamic Random Access Memory), HBM (High Bandwidth Memory), and NAND flash storage.
The dataset is based on John C. McCallum's classic memory-price dataset and is updated monthly for DRAM and NAND, with quarterly updates for HBM.
Users can interact with the data by hovering for details, toggling series, and exporting images. The downloadable CSV file includes every data point along with its source, enhancing transparency and usability.
This comprehensive resource enables engineers and researchers to analyze pricing trends and make informed decisions regarding memory and storage technologies.