TL;DR
Recent analysis shows that memory (HBM) now makes up nearly two-thirds of AI chip component costs, up from 52% earlier this year. This shift impacts supply chain dynamics and chip manufacturing costs.
Memory costs now account for nearly two-thirds (63%) of AI chip component expenses, up from 52% earlier in 2024, according to recent supply chain analysis. This shift underscores changing cost structures in the AI hardware industry, affecting manufacturers and consumers alike.
Analysis based on estimated costs for AI chips from major companies such as Nvidia, AMD, Google, and Amazon shows a significant increase in memory component costs. The data indicates that total component spending on AI chips grew from approximately $22 billion in 2024 to $52 billion in 2025. Memory (HBM) alone contributed around $20 billion of that increase, reflecting a rise in demand and manufacturing costs.
In contrast, other component categories experienced declines or stability: packaging costs fell from 19% to 15%, auxiliary components from 15% to 9%, while logic die costs remained steady at around 13–14%. This trend suggests a reallocation of expenses within the supply chain, with memory becoming the dominant cost driver.
Why It Matters
This development matters because the rising cost share of memory impacts the overall economics of AI hardware production, potentially leading to higher prices for AI chips and related devices. For more on industry challenges, see the price-fixing lawsuit. It also signals shifts in supply chain priorities and may influence future chip design and manufacturing strategies, affecting the pace and cost of AI deployment across industries.
High Bandwidth Memory (HBM) for AI chips
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Background
Prior to this analysis, the industry generally saw memory as a significant but not dominant component of AI chip costs. The recent data, covering quarterly periods from Q1 2024 through Q4 2025, shows a marked increase in memory’s share, driven by supply constraints, demand for higher bandwidth memory (HBM), and increased production costs. The total AI chip market expanded rapidly during this period, with total spending nearly doubling from $22 billion to $52 billion. This growth is partly driven by increased component costs, including memory, as discussed in the lawsuit coverage.
“The surge in memory costs to nearly two-thirds of total component expenses reflects a fundamental shift in the AI hardware supply chain.”
— industry analyst
“The rising demand for high-bandwidth memory in AI chips is driving up costs, which now dominate the component expense landscape.”
— supply chain researcher
AI chip memory modules
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
What Remains Unclear
It is not yet clear whether this trend will stabilize or continue to accelerate. The exact reasons for the rapid increase in memory costs—such as supply chain disruptions or technological constraints—are still being investigated. For context, see how supply chain issues are affecting hardware costs in the price-fixing lawsuit. Additionally, the impact on chip pricing and AI deployment timelines remains uncertain.
GPU memory upgrade kit
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
What’s Next
Industry analysts expect further monitoring of memory cost trends in upcoming quarterly reports. Manufacturers may adjust chip designs or seek alternative memory solutions. Policy and supply chain developments could also influence future costs and market dynamics.
AI hardware memory components
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Key Questions
Why has memory become such a large part of AI chip costs?
Memory, especially high-bandwidth memory (HBM), has seen increased demand due to its critical role in AI processing and data throughput, leading to higher manufacturing costs and a larger cost share in chips.
Will the rising memory costs affect AI chip prices?
Potentially, yes. As memory costs constitute a larger share of total expenses, chip manufacturers may pass some of these costs to consumers, leading to higher prices for AI hardware.
Is this trend expected to continue?
It is uncertain. Market conditions, technological advances, and supply chain factors will influence whether memory costs continue to rise or stabilize in the coming months.
How does this impact AI development and deployment?
Higher component costs could slow down AI hardware deployment or increase the cost of AI services, potentially affecting the pace of AI innovation and adoption across industries.
Source: Hacker News
