Historical memory prices 1960-2026 (dam.stanford.edu)

Article summary

The article presents a graph showing historical memory prices from 1960 to 2026, including DRAM, HBM, and NAND flash prices. The graph displays the price per gigabyte over time, with data collected from various sources. The prices are shown in nominal USD, with an option to adjust for inflation. The graph provides a visual representation of the significant decrease in memory prices over the years.

Main themes

• Historical memory prices
• Memory technology advancements
• Inflation adjustment
• Computing task requirements
• Software development and memory usage

What commenters say

• Adjusting historical memory prices for inflation would significantly alter the graph's appearance, but the logarithmic scale mitigates this effect.
• The notion of pricing memory per gigabyte is not meaningful for older systems, as they did not operate with such quantities.
• The increasing demand for memory is driven by the growing complexity of software and the need for more resources to run contemporary operating systems and applications.
• The recent increase in memory prices may lead to a shift towards more efficient software development and the use of native applications instead of browser-based ones.
• The use of logarithmic scale in the graph helps to accurately represent the significant decrease in memory prices over time, despite the recent plateau.
• The price of memory is still relatively cheap compared to historical highs, especially when considering the cost of other computer components.
• The development of new memory technologies and frameworks may be influenced by the current memory price crunch, potentially leading to more efficient and cost-effective solutions.
• The relationship between memory prices and software development is complex, with some arguing that higher prices will lead to more efficient coding practices, while others believe it will result in the continued use of resource-intensive applications.