When 256GB RAM Costs More Than an RTX 5090: A 3× Price Surge and the World Paying for AI’s Big Bet

Memory prices have officially spiraled out of control.

In just one year, global RAM prices have surged by nearly threefold. One user reported buying a 64GB module for USD 350 only two months ago—today, the same module is listed at USD 2,500. Even more shocking, a 256GB RAM stick now costs USD 1,400 more than an RTX 5090 GPU.

This isn’t speculation or scalping. It’s the result of a full-blown AI-driven global memory crisis.

Reports suggest that AI infrastructure projects led by OpenAI’s Sam Altman have locked up nearly 40% of the world’s DRAM capacity, forcing consumers everywhere to shoulder the cost.

The implications are immediate. Apple reportedly pays around USD 70 per 12GB LPDDR5X chip for the upcoming iPhone 17—up from just USD 25–29 earlier this year. PC makers like Lenovo and HP are rushing to sign pre-purchase agreements with Samsung, SK Hynix, and Micron, as DDR5 contract prices have jumped 123% year-to-date.

Gaming hardware isn’t spared either. Rising memory costs are already pushing up the potential pricing of next-generation consoles.

Memory, once a commodity, is now a strategic global resource.

As manufacturers pivot capacity toward high-margin AI products like HBM, consumer-grade memory is being squeezed out. Analysts warn shortages could persist beyond 2026.

And GPUs may be next. AMD and NVIDIA are expected to raise prices in early 2026, with memory and core costs accounting for nearly 80% of total GPU production expenses. Retailers in Japan have already begun limiting GPU purchases due to uncertain supply.

The industry faces a stark choice: raise prices or cut specifications.

An era of cheap, abundant memory is ending.
And the question remains—are we all ready to pay for AI’s wager?