TL;DR

Consumer RAM prices have jumped far beyond last year’s levels, with 32GB DDR5 kits cited near $375 in early June 2026 and 64GB kits often listed at $600 or more. Industry data and company comments point to AI-driven demand for high-bandwidth memory as the main pressure on DRAM capacity, while the timing and scale of relief remain uncertain.

Consumer RAM prices have surged far above their 2024 and 2025 lows, with a 32GB DDR5 kit cited near $375 in early June 2026 and 64GB kits often listed at $600 or more, as AI demand redirects DRAM capacity toward high-bandwidth memory used in data-center accelerators.

The Thorsten Meyer AI source material cites Tom’s Hardware tracking showing a 32GB DDR5 kit moving from about $80 to $120 a year earlier to $374.97 for the cheapest in-stock kit in early June. A 64GB kit that had been near $150 to $200 through much of 2025 is now described as commonly listing at $600 or more.

Several industry indicators point in the same direction. The source material says DRAM prices rose about 90% in the first quarter of 2026, while HP told investors that memory had reached about 35% of PC build materials, up from 15% to 18% a quarter earlier. HP’s comments mean the price shock is no longer only a retail parts issue; it is feeding into the cost structure of major PC makers.

The reported driver is not a single factory outage or short-term shipping delay. According to the source material, Samsung, SK Hynix, and Micron control most global DRAM production, and the same fabs that can make consumer DDR5 can also make HBM, the stacked memory used next to AI accelerators. HBM reportedly brings far higher revenue per comparable amount of memory, while using roughly three to four times as much wafer area per bit as DDR5.

At a glance
analysisWhen: point-in-time figures from late June 20…
The developmentLate-June 2026 pricing trackers and company disclosures show consumer RAM costs have surged as DRAM makers steer more wafer capacity toward AI-focused high-bandwidth memory.
AI Dispatch · Reality Check · The Memory Squeeze · Part 1 of 10

Why your RAM bill doubled

“Doubled” is the polite version — consumer DRAM is running 3–6× its 2024 lows. The boom-bust cycle that always brought cheap RAM back isn’t coming this time, because the factories that make your RAM now make something far more profitable instead.

The price shock — then vs. now
32GB DDR5 kit$80–120$375
64GB DDR5 kit$150–200$600+
DRAM price move, Q1 2026 alone+90% in one quarter
Memory’s share of a PC’s parts cost15–18%~35%
The mechanism: a zero-sum game inside the fab
1 bit
HBM
=
…of consumer DDR5 wafer area, removed from the world.
One bit of HBM eats 3–4× the wafer area of DDR5. Every wafer shifted to AI doesn’t subtract one wafer of your RAM — it subtracts three or four.
HBM module: $60–100  vs  comparable DDR5: $5–10
HBM now eats ~23% of all DRAM wafer output (up from 19%)
Why it won’t fix itself on the old timeline
~16% supply growth
vs the 20–30% historical norm (IDC, 2026)
Fabs in 2027–28
new capacity is years out; build times in years
~95% in 3 hands
suppliers managing scarcity, not racing to solve it
Locked to 2030
take-or-pay deals spoke for the supply already
The casualties already visible
Micron retired the Crucial consumer brand Apple hiked prices (stock −6%) Framework DDR5 +50% DDR4 now ≥ DDR5 per GB Allocation favors hyperscalers — small buyers last
The take

This is the quiet tax on the whole AI era. Relief isn’t forecast before 2028, and even then prices may settle 30–50% above pre-crisis levels. Buy what you genuinely need now; don’t panic-buy capacity you won’t use. You can’t out-wait the fab math — but, as this series will show, you can shrink what you need. Next: HBM Ate the Fab.

Sources: Tom’s Hardware price tracker; IDC; TrendForce; Counterpoint; Micron Q3 FY26; Wikipedia “2025–present memory shortage”; Sourceability. Figures are point-in-time, late June 2026, and fast-moving.
thorstenmeyerai.com

PC Buyers Absorb AI Costs

The price jump matters because RAM has moved from a cheap upgrade to one of the largest costs in many PC builds. That affects gamers, workstation buyers, small system builders, schools, and businesses replacing aging machines, especially where 32GB has become a practical baseline for modern workloads.

For device makers, higher memory costs can mean higher PC prices, smaller default RAM and storage configurations, or tighter margins. The pressure may hit low-cost laptops hardest because memory makes up a larger share of the bill of materials. Large buyers can sign long-term supply agreements, while smaller buyers may face weaker allocation and more volatile spot pricing.

The broader issue is that consumer hardware is now competing with AI infrastructure spending. If hyperscalers can reserve high-margin memory supply through multi-year deals, the retail PC market may no longer recover on the old boom-and-bust schedule that previously made RAM cheap again after each shortage.

Crucial 32GB DDR5 RAM Kit (2x16GB), 5600MHz (or 5200MHz or 4800MHz) Laptop Memory 262-Pin SODIMM, Compatible with Intel Core and AMD Ryzen 7000, Black - CT2K16G56C46S5

Crucial 32GB DDR5 RAM Kit (2x16GB), 5600MHz (or 5200MHz or 4800MHz) Laptop Memory 262-Pin SODIMM, Compatible with Intel Core and AMD Ryzen 7000, Black – CT2K16G56C46S5

Boosts System Performance: 32GB DDR5 RAM laptop memory kit (2x16GB) that operates at 5600MHz, 5200MHz, or 4800MHz to…

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

HBM Ate The Wafer Supply

High-bandwidth memory is built by stacking multiple DRAM dies and linking them through dense vertical connections. That design gives AI accelerators the bandwidth they need, but the source material says it also makes HBM far less efficient in wafer use than standard DDR5.

That wafer math is central to the squeeze. If one bit of HBM consumes roughly three to four times the wafer area of one bit of DDR5, then shifting capacity toward AI memory removes more consumer supply than a simple one-for-one comparison suggests. The source material says HBM now uses about 23% of total DRAM wafer output, up from 19%.

New capacity is not instant. IDC is cited as expecting about 16% DRAM bit-supply growth in 2026, below the 20% to 30% growth the market has often delivered in past cycles. Micron has also told investors that memory supply may improve gradually in 2028, while saying it does not yet have a clear view of when supply will catch demand.

“Micron said it expects tight industry conditions to persist beyond calendar 2027.”

— Micron earnings materials, reported by MarketWatch

Amazon

64GB DDR5 RAM module

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

Relief Timing Still Unclear

It is not yet clear how high retail RAM prices will go or how quickly any factory additions will lower consumer prices. Forecasts cited in the source material point to no broad relief before 2028, but component pricing can move quickly if demand weakens, new capacity arrives faster, or AI infrastructure spending slows.

The motives of suppliers also require careful wording. The source material says some supply-chain analysis describes manufacturers as managing scarcity, but that is an interpretation of market behavior, not a proven finding of coordinated action. A separate civil antitrust lawsuit reported by Investor’s Business Daily has accused Micron, Samsung, and SK Hynix of price-fixing; those allegations remain claims unless tested in court.

Firepro S9300x2 Standard Airflow

Firepro S9300x2 Standard Airflow

Model: FirePro S9300 x2 – Detailed description of the product's model name

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

Long Contracts Set The Market

The next markers are second-half 2026 contract prices, PC maker pricing decisions for holiday and back-to-school systems, and memory makers’ updates on 2027 and 2028 fab capacity. Micron’s long-term customer agreements through 2030 will also be watched because reserved supply can shape who receives memory first.

For buyers, the practical takeaway is measured. The source material advises buying only the RAM genuinely needed rather than waiting for a rapid return to 2024 pricing or panic-buying capacity that may sit unused. For the industry, the next phase will show whether AI memory demand keeps consumer DRAM expensive even after new fabs begin adding output.

CORSAIR Vengeance RGB DDR5 RAM 32GB (2x16GB) Up to 6000MHz CL36-44-44-96 1.35V Intel XMP 3.0 Computer Memory – Black (CMH32GX5M2E6000C36)

CORSAIR Vengeance RGB DDR5 RAM 32GB (2x16GB) Up to 6000MHz CL36-44-44-96 1.35V Intel XMP 3.0 Computer Memory – Black (CMH32GX5M2E6000C36)

Disclaimer: Maximum Speed requires overclocking/PC BIOS adjustments. Maximum speed and performance depend on system components, including motherboard and…

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

Key Questions

Why did RAM prices rise so sharply in 2026?

The main reported cause is AI-driven demand for HBM, which uses the same broad DRAM manufacturing base as consumer memory but brings higher revenue and consumes more wafer area per bit. That leaves less capacity for DDR5 kits used in PCs.

Are prices really only doubled?

Not in many cases. The source material says consumer DRAM is running three to six times its 2024 and 2025 lows, depending on the kit. The phrase doubled understates the move for some 32GB and 64GB configurations.

When could RAM prices come down?

Current company comments and forecasts cited in the source material do not point to quick relief. Micron has said supply may improve gradually in 2028, but it has not said when supply will fully catch demand.

Should PC builders buy RAM now or wait?

The source material’s guidance is to buy what is genuinely needed now and avoid panic-buying extra capacity. Waiting may help if prices soften, but the reported fab math means a fast return to old lows is not confirmed.

Who is most exposed to the memory squeeze?

PC builders, laptop buyers, small OEMs, and budget device makers face the most visible pressure because they have less leverage than hyperscalers and large manufacturers signing long-term supply deals.

Source: Thorsten Meyer AI

You May Also Like

Just Use Postgres for Durable Workflows

A new approach advocates using Postgres directly for durable workflows, simplifying architecture and leveraging database strengths for scalability, observability, and security.

The Defender’s Window Is Closing Faster Than Anyone Is Counting

Recent developments in AI security reveal rapid advances in offensive capabilities and defensive breakthroughs, raising urgent concerns about future cybersecurity risks.

Quantum Computing revenue surges 9,000% to $3.6M in Q1 2026, but the fine print tells a different story

Quantum Computing Inc. reports a 9,000% revenue increase in Q1 2026 driven by acquisitions, raising questions about organic growth and profitability.

OpenAI co-founder Greg Brockman reportedly takes charge of product strategy

Greg Brockman, co-founder of OpenAI, is now officially overseeing the company’s product strategy, consolidating efforts amid recent leadership changes.