AIThis post was created with the assistance of artificial intelligence (AI).

TL;DR

A new interactive dataset charts memory and storage prices from 1960 to 2026, revealing significant trends in DRAM, NAND, and HBM costs. The data offers insights into industry shifts and future projections.

A comprehensive dataset now publicly available details the evolution of memory and storage prices from 1960 to 2026, offering a long-term view of industry trends. This dataset, built in the spirit of John C. McCallum’s classic memory-price collection, includes interactive visualizations and downloadable raw data, providing valuable insights for industry analysts, investors, and technologists.

Before you orderOffer from Amazon

Get the latest gadgets delivered free with Prime

  • Fast, free delivery on millions of items
  • Prime Video, Amazon Music and more included
  • Member-only deals all year
Start your free Prime trial Free trial for eligible customers · Cancel anytime
As an affiliate, we earn on qualifying purchases.

The dataset covers various memory types, including DRAM, NAND flash, and HBM, with price per gigabyte tracked over the entire period. It highlights the dramatic decline in prices, especially from the 1980s onward, with the lowest recorded prices on a log scale. For example, DRAM prices have fallen significantly across generations, from pre-DDR to DDR5, reflecting technological advancements and manufacturing efficiencies.

Additionally, the dataset models cost breakdowns for AI accelerators, including components like HBM memory, logic die, and packaging, based on estimates from Epoch AI. It also provides projections for HBM prices by generation, with HBM4 expected to launch in Q3 2026. Learn more about memory pricing trends on our homepage. These estimates are derived from industry analyst reports, as HBM is sold under confidential contracts with no public spot market.

At a glance
reportWhen: published March 2026
The developmentAn extensive dataset compiled from historical sources and industry estimates illustrates how memory prices have changed from 1960 through 2026, providing a long-term view of industry trends.

Implications of Long-Term Memory Price Trends

This dataset is significant because it offers a detailed view of how memory technology costs have evolved over more than six decades, informing industry, investment, and research decisions. The decline in prices has enabled exponential growth in computing power, especially for AI and data centers. Understanding these trends helps predict future price movements and technological shifts, influencing hardware design and market strategies.

Products Worth Considering

As an affiliate, we earn on qualifying purchases.

Historical Development of Memory Pricing

Memory prices have historically fluctuated due to technological breakthroughs, manufacturing scale, and market demand. The 1980s marked the beginning of rapid price declines, driven by the adoption of more efficient manufacturing processes and new memory generations like DDR. Recent years have seen prices stabilize at lower levels, with continued reductions in HBM costs projected through 2026. Prior to this dataset, detailed long-term price tracking was limited, making this compilation a valuable resource for understanding industry evolution.

“The long-term data shows a clear downward trend in memory prices, which has been fundamental to the growth of modern computing.”

— an anonymous researcher

Limitations of Historical Price Data and Projections

While the dataset provides a comprehensive view, some figures, especially for earlier decades, rely on estimates and approximations. HBM prices are modeled estimates based on industry analyst reports, as no public spot market exists for this memory type. Future projections, such as HBM4 pricing, are based on industry forecasts and may be subject to change due to market dynamics or technological breakthroughs.

Future Updates and Industry Impact of Memory Pricing Data

The dataset will be periodically updated to include new pricing data and refined estimates. Industry analysts and hardware manufacturers will likely use this information to inform product development, pricing strategies, and investment decisions. Further research may explore how emerging memory technologies could alter long-term cost trends and industry competitiveness.

Key Questions

How accurate are the historical memory prices in the dataset?

The prices are based on the cheapest retail listings and industry estimates, especially for earlier decades. While they provide a reliable trend overview, some figures are approximations and may not reflect exact transaction prices.

Why is HBM pricing modeled rather than directly measured?

HBM is sold under confidential contracts with no public spot market, so industry analysts estimate prices based on component costs and bandwidth metrics, making the figures modeled estimates rather than direct transactions.

What does this dataset reveal about future memory costs?

The data indicates a continuing decline in memory prices, particularly for HBM, with projections suggesting further reductions in the coming years, which could lower costs for AI accelerators and data centers.

How can this data influence industry decisions?

Understanding long-term price trends helps companies plan hardware development, optimize supply chains, and strategize investments in emerging memory technologies.

Will the dataset include future projections beyond 2026?

Future updates may include projections based on industry forecasts, but precise long-term predictions remain uncertain due to market and technological variables.

Source: Hacker News

You May Also Like

Ultrawide Monitors Explained Without the Hype

The truth about ultrawide monitors reveals how they can enhance your workspace and gaming experience, but understanding their real benefits is essential before making a choice.

What To Look For In The 9 Best E Ink Tablets For Reading And Handwriting In 2027

A nine-device buying guide compares screen size, color, handwriting features and software options in E Ink tablets for 2027.

Why not yserver? It’s Xserver, but Rust-y.

Yserver, a new display server written in Rust, aims to provide a modern alternative to Xorg, supporting Linux desktops with Vulkan and limited multi-monitor support.

Show HN: A Project Oberon System Version Running On RISC-V Instead Of RISC-5

A developer has shared a version of the Project Oberon operating system running on RISC-V, replacing the original RISC-5 platform, marking a significant porting effort.