DRAM prices have surged—reports cite export prices rising as much as 497%—because AI data centers require huge volumes of memory, while manufacturers are shifting production toward high‑bandwidth memory for AI chips,... Three companies—Samsung Electronics, SK hynix, and Micron—control roughly 95% of global DRAM supp...

Create a landscape editorial hero image for this Studio Global article: What is driving the reported 497% surge in DRAM prices, how is AI data-center demand from major cloud companies straining memory-chip supply. Article summary: The reported 497% DRAM price surge appears to be driven mainly by AI infrastructure demand: hyperscalers are buying large volumes of high-bandwidth memory and server DRAM, while memory makers are shifting capacity away f. Topic tags: general, general web. Reference image context from search candidates: Reference image 1: visual subject "The surge is being driven by unprecedented demand from major AI players such as Nvidia, Google, and AMD, whose data center GPUs and AI" source context "Digital IT News, Information Technology News, Tech News, Computer news, breaking news , analysis news" Reference image 2: visual subject "The surge is being driven by unprecedent
The global boom in artificial intelligence infrastructure is creating an unexpected bottleneck: memory chips. Demand for DRAM—the primary working memory used in servers, PCs, and many electronic devices—has surged as cloud companies build massive AI data centers. The result has been rapidly rising prices and a tightening supply chain across the semiconductor industry.
Some industry reports say DRAM export prices from South Korea have increased as much as 497% over the past year, highlighting how sharply demand has risen alongside the expansion of AI infrastructure. [1]
The biggest driver is the explosion of AI workloads. Training and running modern AI models requires enormous amounts of memory bandwidth and capacity, far more than traditional computing tasks. As a result, AI servers rely heavily on advanced memory technologies such as high‑bandwidth memory (HBM), which is built from specialized DRAM. [9]
To meet this demand, memory manufacturers have shifted production capacity toward HBM and server‑grade DRAM. Because these products are more profitable and critical for AI systems, they take priority over commodity memory used in PCs and smartphones. [9]
This shift has tightened supply for the rest of the market. When the same fabrication lines must serve both traditional DRAM and AI‑focused memory, allocating more capacity to HBM inevitably reduces availability of standard chips—pushing prices higher.
Studio Global AI
Use this topic as a starting point for a fresh source-backed answer, then compare citations before you share it.
DRAM prices have surged—reports cite export prices rising as much as 497%—because AI data centers require huge volumes of memory, while manufacturers are shifting production toward high‑bandwidth memory for AI chips,...
DRAM prices have surged—reports cite export prices rising as much as 497%—because AI data centers require huge volumes of memory, while manufacturers are shifting production toward high‑bandwidth memory for AI chips,... Three companies—Samsung Electronics, SK hynix, and Micron—control roughly 95% of global DRAM supply, giving them major influence over pricing during the AI driven memory boom.
Hyperscalers such as Microsoft and Google are locking in huge memory purchases for AI servers, causing suppliers to prioritize data‑center customers over consumer electronics.
Continue with "Why Ant Group’s Profit Plunged 79% — and What Its AI Bet Means for the Future" for another angle and extra citations.
Open related pageCross-check this answer against "Starship Flight 12: What to Know About SpaceX’s First V3 Launch From Starbase Pad 2".
Open related page- Micron Technology (MU), Samsung Electronics, and SK Hynix collectively control 95% of global DRAM supply, with Micron’s revenue tripling year-over-year to $23.9 billion and adjusted earnings climbing eightfold as HBM demand surges; DRAM export prices from...
Samsung predicts that memory supply shortages will worsen further in 2027, and customers have already started placing orders for 2027. Amid a continuing global shortage of semiconductors (memory) , semiconductor giant Samsung has warned that 'the global mem...
Samsung Electronics and SK hynix, according to Hankyung, have reportedly pitched first-quarter server DRAM prices up 60–70% from Q4 last year to major clients including Microsoft and Google, while PC and smartphone DRAM buyers are seeing comparable hikes. T...
The price pressure is also amplified by the sheer scale of global AI investment. Major technology companies are collectively spending hundreds of billions of dollars on data‑center infrastructure, creating a surge in demand that memory supply has struggled to match. [1]
Hyperscale cloud providers—including companies like Microsoft and Google—are among the biggest buyers of server DRAM for AI clusters. Reports indicate that suppliers such as Samsung and SK hynix raised server DRAM prices by around 60–70% in a single quarter for large customers as demand intensified. [7]
At the same time, manufacturers are prioritizing AI infrastructure buyers over lower‑margin markets. Data‑center operators and AI server builders are absorbing a large share of global memory output, leaving PC and smartphone manufacturers competing for a smaller pool of chips. [11]
Another constraint is manufacturing complexity. Producing HBM requires advanced packaging techniques and lengthy qualification cycles, which limits how quickly companies can increase supply. Even if chipmakers invest heavily in new capacity, it takes time for additional production to come online. [9]
Unlike many technology sectors, the DRAM market is extremely concentrated. Just three companies dominate global production:
Together, these firms control roughly 95% of worldwide DRAM supply, giving them significant influence over prices and production priorities. [1]
Because the same companies also lead the market for HBM—an essential component in AI accelerators—the AI boom has strengthened their pricing power even further. [9]
Industry forecasts suggest that the current tight supply conditions will not resolve quickly.
Several analysts expect DRAM supply constraints and higher prices to continue through at least 2026, driven by sustained demand from AI infrastructure projects. [10][
14]
Some industry leaders are even more cautious. Samsung has indicated that global memory shortages could persist into 2027, with customers already placing orders years in advance to secure supply. [3]
Analysts also expect DRAM prices to keep rising in the near term. For example, some forecasts suggest average DRAM selling prices could increase by about 33% in 2026, reflecting the ongoing imbalance between supply and AI‑driven demand. [10]
The current surge in DRAM prices is not just another semiconductor cycle. The rise of AI computing has fundamentally changed how memory is consumed, with data centers becoming the dominant source of demand.
As hyperscale companies continue expanding AI infrastructure—and as advanced memory like HBM becomes essential for high‑performance GPUs—the pressure on DRAM supply may remain a defining feature of the semiconductor market for several years. For businesses and consumers alike, that likely means higher memory prices and tighter supply until production capacity catches up with the AI boom.
As leading memory semiconductor manufacturers; Samsung Electronics Co. Ltd. (KRX: 005930), SK hynix Inc. (KRX: 000660), and Micron Technology Inc. (NASDAQ: MU), divert production capacity toward high-bandwidth memory, a premium form of DRAM used in AI data...
As things currently stand, the AI industry's gargantuan demand for these chips will keep memory prices elevated, possibly for years. According to Bank of America analysts cited by SK Hynix, the average selling price for DRAM will increase by as much as 33 p...
Memory prices are running hot again, with the latest forecasts pointing to steep quarter-on-quarter increases for both DRAM and NAND in early 2026. The dynamic is familiar: Hyperscalers and AI-server builders are soaking up supply, while manufacturers prior...
The global memory market is entering an unprecedented supply-constrained phase as soaring demand from artificial intelligence (AI) infrastructure collides with strategic capacity reallocations by leading semiconductor manufacturers. According to the latest...