Memory can account for almost 60% of the materials cost in handsets priced around $400, Fairphone Chief Executive Raymond van Eck said in interviews published September 16. The estimate concerns the bill of materials, the parts a manufacturer buys to build a phone, rather than the handset’s retail price. We could not independently verify van Eck’s estimate. For smaller manufacturers, the immediate problem is securing an allocation of chips at all, not merely paying the quoted price.

“If you don’t have allocation, you’re out of the game anyhow,” van Eck said. Fairphone has not raised its phone prices, even as the company says tighter memory supply has pushed up its component costs. In a September 3 company note, van Eck linked its access to multiyear supplier relationships.

Memory suppliers prioritize AI and server applications when allocating production capacity, reducing the supply available for PC and phone memory.

Key Takeaways

AI-generated summary, reviewed by an editor. More on our AI guidelines.

Two boards for one phone

Finnish phone maker Jolla designed two motherboard layouts so it can use either a combined memory-and-storage package or separate chips, depending on what suppliers can deliver. Its batch sales model also lets the company set memory configurations for limited production runs. The September II batch lists 8GB of RAM and 128GB of storage, with a paid 12GB and 256GB option.

Jolla also stress-tests samples from every incoming batch. The checks are meant to determine whether chips sold as new are actually refurbished parts being resold as first-hand.

The company’s component costs have not moved in a straight line. Its combined storage and DRAM package reached a peak at the end of March and had held at that level by September 16. That plateau ran against spring forecasts that the package price would double by autumn.

Orders without final terms

Framework, the repairable-laptop maker, places non-cancellable memory orders far in advance without knowing the final price, delivery timing or volume. Chief Executive Nirav Patel said the supply mismatch became clear late in 2025, setting off “everyone trying to grab as much supply, as much inventory as they could, as quickly as they could.”

The company’s modular design gives buyers another option: they can install compatible memory from an older laptop or buy used modules. Framework still had to change configurations and prices in July after its supplier quoted LPCAMM2 costs at more than twice the level of its prior inventory.

On September 8, Framework said it had secured a limited quantity of lower-cost Micron 32GB and 64GB modules. It reduced prices for some pending orders and issued refunds on eligible shipped orders. The new prices remained above Framework’s original preorder baseline, and access was initially limited to existing customers through Batch 10.

Know someone who'd find this useful? ✉️ Email it to a friend in one click, or they can subscribe free here.

Prices rise unevenly

A September 3 model-level tracker covering more than 4,000 active smartphones found that retail prices for existing models had risen 15% on average during 2026. More than 40% of models had received price increases, while new launches cost 25% more than the models they replaced a year earlier.

The increases varied by market. Prices rose 21% in India during 2026, compared with 7% in Europe and 5% in the United States. Manufacturers were also reducing storage, simplifying camera configurations and bringing back some 4G models to contain costs.

A July 3 pricing survey projected conventional DRAM contract prices would rise 13% to 18% quarter-over-quarter in the third quarter of 2026. Weaker consumer demand was limiting how much more phone makers would accept.

Jolla’s chief executive, Sami Pienimäki, expects supply to normalize only from 2028. For now, he said of the plateau in its package costs, “We are very happy for that, of course, for the time being.”

Frequently Asked Questions

Does memory cost 60% of a phone’s retail price?

No. Fairphone CEO Raymond van Eck’s estimate concerns the bill of materials, meaning the parts used to build a phone priced around $400. The estimate has not been independently verified by Implicator.

How is Jolla adapting to the memory shortage?

Jolla designed two motherboard layouts to accommodate combined packages or separate chips. It sells in batches with different memory configurations and stress-tests incoming samples for refurbished chips resold as new.

Why does Framework order memory before knowing the final terms?

Securing supply is the immediate constraint. Framework places non-cancellable orders well ahead without knowing the final price, delivery timing or volume. Its modular laptops also let customers reuse compatible memory.

Are memory prices rising everywhere?

The pressure is uneven. Jolla’s combined storage and DRAM package reached a peak in March and held through September 16. Framework announced limited lower-cost inventory on September 8, enabling reductions and refunds while prices remained above its original preorder baseline.

How does AI demand affect phone and laptop memory?

Memory suppliers prioritize AI and server applications when allocating production capacity, reducing the supply available for PC and phone memory. Jolla’s CEO expects supply to normalize only from 2028.

AI-generated summary, reviewed by an editor. More on our AI guidelines.

Nvidia Sets October Launch for RTX Spark N1X Without Naming a Price
Nvidia set an October 2026 shipping window for RTX Spark computers. The RTX Spark lineup will have two N1X versions, with unified memory topping out at 128GB. With sales due next month, no vendor has
RAM Prices May Ease in 2028. High-VRAM GPUs Are a Different Story
When Allan Witt loaded a local AI model, the hardware specialist needed more graphics memory than Nvidia's top consumer card, the 32 GB RTX 5090, provides. Witt, co-founder and editor in chief of Hard
Amazon Hikes Device Prices 60% While Spending $220 Billion on AI
In late July 2026, Amazon Chief Executive Andy Jassy told investors the company’s 2026 capital spending would reach $220 billion, up from a previous estimate of $200 billion, with higher memory costs
AI News

San Francisco

Editor-in-Chief and founder of Implicator.ai. Former ARD correspondent and senior broadcast journalist with 10+ years covering tech. Writes daily briefings on policy and market developments. Based in San Francisco. E-mail: editor@implicator.ai