AI Demand Triggers Memory Crisis: DRAM Prices Surge 50% as Supply Chain Buckles

Server DRAM prices surge up to 50% as AI-induced memory shortage hits hyperscaler supply -- U.S. and Chinese customers only getting 70% order fulfillment

U.S. and China buyers now see just 70% order fulfillment, as AI demand drives DDR5 shortages and PC parts follow suit. The DRAM supply chain is choking, and server memory is taking the first hit. According to DigiTimes, major U.S. and Chinese hyperscalers are now receiving just 70% of the server DRAM they order. That's despite agreeing to contract price increases of up to 50% for Q4, well above the 30% hike many buyers had budgeted for earlier this year. Naturally, AI sits at the core of all this. While it's HBM that gets the headlines, demand for conventional DDR5 RDIMMs is also outpacing supply, particularly at advanced nodes where Samsung and SK hynix have diverted capacity toward parts bound for AI acceleration. Samsung's recent pricing adjustments confirm the reprioritization, with the company having raised server SSD prices by up to 35% and RDIMM contract rates by as much as 50%, citing sustained demand from enterprise and cloud customers. What's left is a market where even the largest buyers can't secure enough memory. Spot prices have surged since late September, and several top-tier suppliers are reportedly refusing to quote for October allocations. DDR5 16 GB modules that traded at $7 to $8 last month are now hovering around $13, with availability tightening further into November. Module makers are bracing for out-of-stock situations by the end of the quarter. The rest of the market fares worse. Channel players and smaller OEMs are seeing order fulfillment rates closer to 35% to 40%. With hyperscalers locking in fixed allocations, lower-priority customers are pushed to the spot market or told to wait until capacity opens up in 2026. Micron warned of this in its most recent earnings call, telling investors that DRAM is a "tight industry" and that bit supply growth will lag demand through the end of next year. TrendForce also flagged a potential quote freeze across certain modules, as suppliers shift to day-to-day pricing in China and avoid locking themselves into bad deals. This is causing retail DDR5 prices to creep upward, with no sign of stabilization before year's end. DDR4, meanwhile, is in slow decline. China's Nanya Technology recently said standard DDR4 makes up just 20% of the total DRAM market, and it's no longer being prioritized for volume production. Unless demand unexpectedly cools or yields improve sharply, DRAM allocation for everyone outside the top buyers will be constrained into 2026.

'This Is Insanity': DDR RAM Prices Soar Due to AI Demand

When he's not battling bugs and robots in Helldivers 2, Michael is reporting on AI, satellites, cybersecurity, PCs, and tech policy. Don't miss out on our latest stories. Add PCMag as a preferred source on Google. Yikes: If you're building a PC or want to add more memory, be prepared for sky-high pricing on DDR4 or DDR5 RAM. RAM pricing has soared in recent weeks to shocking levels. "This is insanity," wrote one Reddit user, who reported buying 32GB of DDR5 RAM for $104 only to see the price shoot up to $220. I bought a two-pack of 16GB DDR4 RAM sticks in May from Amazon for $51.99. The price has since jumped to an eye-watering $114.99. PCPartPicker, which monitors consumer RAM pricing at retailers, has also noticed the upward trend; for DDR5 RAM, the price increases started in September before accelerating this month. What's going on? Research firm TrendForce, which tracks the memory market, reports demand for AI, including the construction of new data centers, has contributed to a memory shortage. This includes some tech companies stockpiling DRAM while signing longer-term deals with memory suppliers. To meet the AI demand, the two top memory makers, Samsung and SK Hynix, have also been prioritizing HBM (high bandwidth memory) DRAM, according to Korea's The Chosun Daily, which notes: "From a supply perspective, production capacity for general-purpose DRAM is inevitably shrinking." As for DDR4 RAM, pricing began to shoot up in June amid reports that memory makers including Samsung and Micron stopped producing the older RAM to focus on DDR5 and HBM RAM. TrendForce adds that some suppliers are "aggressively" shifting capacity toward server-grade memory, constraining supplies for PC-focused DDR5 and DDR4 supplies. "As a result, PC DRAM prices are forecast to continue edging upward through the quarter," the research firm wrote last month. Historically, DRAM pricing has been cyclical, with ups and downs involving memory makers ramping up production, overproducing and then cutting manufacturing as demand falls. So the high prices aren't exactly a surprise. Still, it looks like the AI rush risks creating a longer, persistent shortage for memory. In the meantime, some PC builders plan on waiting in the hope retailers will offer discounts next month during Black Friday sales.

Hold onto your wallets - PC building is about to get even more expensive

PC building was never a cheap hobby. From the very start, owning, maintaining, and upgrading computers has been a somewhat costly affair, and while it has certainly become more accessible in the 21st century, its cost has begun to creep upward, even when you adjust for inflation. There have always been cycles of affordability and scarcity, but AI (and the demand for it) has created a new kind of demand shock. The big players in the AI space, like OpenAI, Microsoft, and Google have been snapping up large swaths of silicon, which is required for the many different components that LLMs and other AI workloads use in massive quantities. Previously, sudden rises in cost for PC hardware were attributed to temporary phenomena, like the crypto mining boom and a worldwide pandemic, but AI's current hunger for memory and compute could keep PC hardware prices inflated for years to come. PC hardware has slowly become more expensive Cost creep is very real Even before the AI boom, the cost of entry for enthusiasts had been steadily climbing. Component prices began rising after the pandemic disrupted logistics and manufacturing capacity. Silicon nodes shrank, complexity increased, and those costs were passed along to consumers. That isn't a dig at manufacturers either; we're talking about components that are incredibly complex, and when new standards debut, of course they'll be more expensive. DDR5 was not cheap when it launched, but like most types of memory, as more of it is put into circulation, prices come down, allowing more builders to have access to it. The early adopter tax is certainly real, but so are shipping costs and aggressive product segmentation, which all the big players are guilty of to some extent. Those factors have kept prices rising, and it's about to get even worse. AI workloads are driving up hardware costs We're seeing the effects now LLMs like GPT-5, Gemini, Claude -- you name it, require immense amounts of DRAM and compute throughput to function. Each model contains billions of parameters and an unfathomable amount of training data that is loaded into high density memory. That memory isn't so different from the kind you and I put in our PCs. Samsung and SK Hynix, two of the biggest memory manufacturers in the world, have both entered a partnership with OpenAI, and have agreed to produce 900,000 DRAM wafer starts per month at an accelerated capacity. This could have massive consequences for the PC hardware space, as these silicon wafer starts produced by Samsung and SK Hynix are the same type that are used in things like CPUs and GPUs in addition to memory. The effects of this are already being felt. According to data from PCPartPicker, DDR5 prices have gone up by anywhere between 30-50% in the last month or so following this announcement. 32 GB kits of DDR5-6000 jumped from around $125 to well over $200 at the time of writing. Other specs follow the same price increase, with DDR5-5600, DDR5-5400 and even DDR5-4800 rising similarly. DDR4 isn't safe either, but it's possible that the increase in price is due to the fact that production of DDR4 is beginning to slow as the demand for it decreases. Short-term price increases like these can also be attributed to things like tariffs, but it's yet to be seen how those will affect prices in the long-term. We've seen this before Well...sort of Silicon scarcity isn't anything new to PC builders who lived through the crypto mining boom of the early 2020s, where the worldwide stock of graphics cards was completely engulfed by those looking to strike digital pay dirt. Prices skyrocketed, bots flipped inventory, and true enthusiasts were left salvaging cards from pre-builts, or paying double (or even triple) MSRP for individual cards. But eventually the craze came to a close, and stock mostly returned to shelves. Even the scarcity experienced during the most recent GPU launch cycle eventually ended, but a diversion of supply to AI workloads could cause a much longer-term scarcity that I don't think most enthusiasts are ready for. The future of affordable PC hardware looks grim AI is so much more lucrative for manufacturers Companies like OpenAI plan to build data centers at an incredibly rapid rate, and the trickle-down effect on regular consumers doesn't look like a positive one. When major buyers corner the market for raw silicon, smaller segments (which includes DIY PC building) are left with less supply, and therefore higher costs. Fabs can't instantly scale to meet this new demand, and even small bottlenecks ripple through the market, as we've seen in the past. A single-digit percentage shift in wafer allocation toward AI servers can mean millions fewer DDR5 chips available for consumers, let alone a double-digit percentage. On one hand, I understand it. Nvidia's evolution from "graphics card company" to "AI behemoth" has made the gaming market largely an afterthought for them; they make several billion dollars more from compute than they do from graphics, so why would they do anything other than strike $100 billion deals with AI giants? I hope I'm wrong The crypto boom of the early 2020s showed how a niche, high-demand market can wreak havoc on PC part pricing. But unlike mining, AI isn't going away. It's not a fad or a speculative bubble, but a structural shift in how silicon is consumed. Every year, models get larger, inference scales wider, and memory requirements climb higher, and unfortunately, I think that spells bad news for us PC hardware enthusiasts, but I'd love to be wrong.

AI-driven demand pushes DRAM prices up 50%, orders go partially unfilled

Serving tech enthusiasts for over 25 years. TechSpot means tech analysis and advice you can trust. The big picture: The AI boom is driving DRAM demand to record highs, and the market is struggling to keep up. According to a new report from DigiTimes, leading cloud service providers in the US and China have agreed to pay top dollar on orders that are only being partially filled. In September, an anonymous source outlined Samsung's plans to raise DRAM and NAND flash prices in the fourth quarter of 2025. It is now late October, and the situation has seemingly played out much like the source said it would. Per DigiTimes, some major hyperscalers agreed to fourth quarter contract price increases of up to 50 percent over what they'd been previously paying for DRAM - an increase of 30 percent more than most expected to be paying in Q4. Given the markup, you would expect prompt delivery but that apparently has not been the case. According to the publication, buyers are only receiving around 70 percent of the server memory they ordered, and the situation could get worse before it gets better. Tom's Hardware notes that memory makers are preparing for out of stock situations by the end of the year. Should that come to fruition, the trickle-down effect could force some hyperscalers to slow their AI expansion efforts. Keep in mind that this only applies to major cloud service providers with deep pockets. For smaller players and OEMs, the situation is even worse. Tom's notes that order fulfillment rates for lower-priority customers is closer to 35-40 percent. To meet demand, smaller players will likely end up having to pay even more or sit on the sidelines and wait until supply improves. On the retail side, not much has changed as it relates to DDR5 versus DDR4. The older tech is being phased out, resulting in higher spot prices compared to the newer tech. A quick check of DRAM Exchange reveals a daily high of $28.00 for DDR4 16Gb (2Gx8) versus $20.00 for DDR5 16G (2Gx8). Image credit: Charlie Belvin, Athena Sandrini

RAM and storage is ridiculously expensive right now because of *drumroll* AI, of course, and there's little reason to think prices will drop any time soon

Consumer storage isn't doing bad, but that will probably change before long. We can't seem to get a moment's respite in the PC gaming industry, can we? We spent the better part of this year lamenting ridiculously high GPU prices, and as soon as said prices finally start dropping, memory and storage prices start climbing. You might have noticed this worrying trend if you've been looking for a RAM upgrade lately. Charts on pcpartpicker.com are telling: since mid-2025, DDR4 and DDR5 consumer memory kit prices have been climbing, with particularly notable spikes since September. That's true all the way from DDR4-3200 kits up to DDR5-6000 ones. To take one concrete example, this 32 GB Corsair Vengeance DDR5-6000 kit on Amazon, which was stable at $95 or less throughout most of the year, but from mid-August onward started climbing. Now, it's up to $184, which makes for a 94% increase. This, after a good few months of sustained DDR5 price drops through the tail end of 2024 and first half of 2025. DDR4 isn't safe, either, as this 32 GB Corsair Vengeance DDR4-3600 kit on Amazon shows. It was stable at about $70 through most of the year, but excluding a brief drop back to down to this original price, from mid-August it shot up to a new price of $161. That's a 115% increase: over twice its previous cost. This trend fits all the market analysis we're seeing lately, which by now seems little more than pointing out the obvious: memory is becoming incredibly expensive. The "premium" you have traditionally had to pay for DDR5 compared to DDR4, according to a recent chart spotlighted by a Futurum analyst, no longer exists. In fact, in just the space of one week, spot prices for DDR4 memory recently doubled. DDR4 memory is winding down production, which means less supply on the market. But we wouldn't necessarily expect this factor to cause such dramatically rising prices on its own. The fact that DDR4 prices are rising by such an extreme amount indicates that there simply isn't enough supply compared to demand across the memory industry in general, and that companies are looking for whatever memory they can get their hands on. This tracks with what's been going on in the DRAM and memory industry. Samsung and SK Hynix, for instance, are reportedly charging customers up to 30% more for DRAM and NAND in the fourth quarter of 2025. Note that that particular price hike includes NAND flash memory, too, which is what goes into non-volatile SSD storage. We've not seen as much of a reflection of rising flash storage prices at the consumer level as we have with RAM just yet, but it could be coming. Adata has said that supplies of DRAM, NAND, and even HDD storage are now in shortage, for the first time in 30 years. Many of the best SSDs for gaming are still pretty cheap, but I'd bet it won't be long before these shortages start showing effects in the consumer market, just like has started to happen with memory. It might not be long before both RAM and SSDs are prohibitively expensive. The main cause of these rising memory and storage prices is of course the tech industry's summer sweetheart, artificial intelligence. In addition to giving us plenty of slop content to whine about, AI is also making our PC gaming hardware more expensive. AI servers tend to require an astronomical amount of storage and memory, and the processors have to be able to access data in said storage and memory pretty quickly. Training AI requires great volumes of data, as does running inference once the AI is trained -- and often re-training is continually ongoing. Usually this data is unstructured, as AI will pull from lots of different areas as it learns and forms its neural network or probabilistic models. SSDs are ideal for this when it comes to storage, and for memory, high bandwidth memory (HBM) is what most AI server companies are looking for. This isn't the same as the DDR memory we get in our gaming PCs, but the memory chips are DRAM just like we use in our familiar DIMMs. AI servers buying up HBM means less DRAM in the market for our RAM kits, which causes prices to rise. The same can be said for SSDs, if a little less directly. Even if the SSDs in AI servers aren't the same as the ones we put in our gaming PCs, NAND production for these AI server SSDs can detract from NAND production for consumer SSDs once supply is limited. Usually, AI servers use higher-quality SSDs than consumer ones, with different NAND, but when storage gets tight, AI server customers can start ordering more consumer-oriented NAND. In fact, we've seen hints of this already, as Trendforce's investigations showed that cloud providers are increasingly being pushed towards QLC ('quad-level cell') SSDs like the ones found in many consumer desktops and laptops. This is driven in part because nearline HDDs (ones for infrequently accessed storage) are in short supply. Yep, that's right, even traditional hard drives are facing shortages and, surprise surprise, price hikes. Western Digital, for instance, recently raised prices for all its HDD products. Memory shortages and rising prices have been predicted for a while now -- over a year ago, for instance, Micron said that the AI would "impact the supply demand balance" as it's "made memory sexy again". Storage shortages, not so much. Last year, Trendforce analysis predicted that storage prices would decline by as much as 15% over the final quarter of 2024, and there was no prediction of the shortages we're seeing now. Peaks and troughs are normal in the memory industry, due to the ebbs and flows of supply and demand, but the recent shortages and price hikes don't fit that mould. They're simply too extreme. What explains these prices is the unprecedented scale of investment in AI. Over the last few months, we've seen the AI industry receive and shuffle around some seriously mind-boggling amounts of cash and capital. OpenAI's Stargate project was announced at the start of the year and plans to spend $500 billion on AI infrastructure in the US over four years. That project is now thoroughly underway, and has even spread beyond the borders of the US into the UK. We've seen Nvidia recently invest $100 billion into the Sam Altman-headed company, and OpenAI invest tens of millions into AMD and agree to purchase $300 billion of computing power from Oracle over the next five years. Plus many more deals and investments, such as with Broadcom -- and these are just considering OpenAI, the poster child for this AI era alongside Nvidia. With all this money being flung around and data centers springing up so quickly that they're having to be housed in tents, it's little wonder that the memory and storage industries are experiencing a drain in supply. Which is great for them, of course, as they can rake in the money from raised product prices and booming share prices, as well as sell out their entire HBM supply, but it's not so great for us. There's little chance that these memory and storage manufacturers will create much more supply than is currently needed, either. For one, this can sting these companies if demand unexpectedly drops, as they discovered in 2023. Second, whirring up new or increased semiconductor production is no simple matter and can take years, and the demand from the AI sector doesn't seem to be slowing down. There is, of course, the chance that the entire AI industry gives a loud and audible pop as what many are calling the AI bubble bursts. To simplify, the reason some think the AI industry is a bubble is because there's tons more investment than revenue in the industry right now, which arguably means more and more money must be pumped into it to keep it going under the hopes that some day the revenue will appear. To this end, a lot of the money involved seems to be flowing back and forth between the same companies. For instance, Nvidia invests in OpenAI, and OpenAI buys hardware from Nvidia. This creates what some see as a circular system where money is continually injected with little actual return, inflating the metaphorical bubble until eventually it pops like the housing market did in 2008 and the dot-com market did in 2000. We must remember, however, that although the dot-com bubble burst, it did leave behind the seeds for the internet. In a similar way, the AI bubble bursting could leave behind the seeds for a less inflated world of AI. That could still sort out the memory and storage market, though. As it stands, vast swathes of memory and storage capacity is spoken for by big AI companies well before it's packaged and ready to be installed into servers, most of it in advance. And these production commitments are based on the current rate of arguably inflated AI market expansion. If that inflation comes to a sudden end and reversal, we might be dealing with oversupply again and lower prices. However, market analysis sees the current memory shortage lasting for another three or four years. This tracks with at least what some, such as ex-Intel CEO Pat Gelsinger, think about the so-called AI bubble: that sure, it might burst, but that won't be for "several years." In which case, we might have to settle in for a long haul of high memory and storage prices.

In case we needed a visceral reminder just how screwed memory prices are thanks to AI, Samsung and SK hynix are apparently raising prices by up to 30%

I don't own a crystal ball, but if I did, I'm sure it would be telling me all about memory prices continuing to climb. No doubt those will be the words tumbling along the semiconductor industry winds for some time. The latest on this front, in case we needed any more evidence, is that according to the Korean Economic Daily (via TrendForce), "major memory suppliers such as Samsung Electronics and SK hynix are [charging] customers up to 30% higher prices for DRAM and NAND flash in the fourth quarter of this year" (machine translated). The "semiconductor industry" says that the memory supply shortage will last for three or four years, which, of course, means increased prices thanks to supply and demand. The same sources apparently predict that the high bandwidth memory (HBM) market will grow to $100 billion by 2030. This doesn't seem too far-fetched. Recently, the head of Hyundai Motor Securities Center said that "the HBM market size next year will exceed 50 billion dollars" (via Jukan on X). And with the current skyrocketing direction that RAM pricing is heading, even from a layman's perspective, this tracks. These price rises are caused by -- can you guess? -- AI, of course. Big AI companies like OpenAI seem to be signing big contract after big contract, which means the need for tons of DRAM for HBM and NAND for storage to go into all the newly contracted datacentres. Even countries, such as the UK, are signing investment partnerships with companies such as Nvidia, OpenAI and Microsoft to get the ball rolling on AI infrastructure. It's not just the latest memory technologies being guzzled up, either. We're seeing DDR4 memory rise to the point where it's barely cheaper than DDR5, even as DDR5 prices rise. And at the other end of the spectrum, even HDDs are facing supply shortages. We've heard companies sounding the alarm about these supply shortages for a while now, so none of this is really a surprise, but seeing big numbers like 30% really does bat the message home. It can be tempting to think of all this as affecting just the AI industry, but memory is memory: it all comes from the same manufacturers and factories, whether it ends up as HBM in a datacentre or as part of a gaming PC's dual RAM kit. So unfortunately, it's likely these price hikes for big AI customers will also trickle down into consumer prices for desktop RAM, and indeed, we're already seeing as much. Given the long timeframes analysts and business executives seem to be discussing, there might be no better time than now to pick up some RAM if you're thinking of upgrading. Although I suppose Black Friday is just around the corner -- maybe that'll be the time.

AI has ruined your Christmas this year if you were hoping to find DDR5 RAM under the tree

There are already a lot of reasons to hate the AI boom. Take your pick, is it the detrimental and possibly irreversible impacts it's having on the environment? Is it the massive increase in data farming that's happening to train all of the AI models? If it's neither of those, I'm about to give gaming PC enthusiasts a whole new reason to hate AI. It's taken years for DDR5 RAM to come down in price to the point where gamers feel it's actually worth upgrading to. Yes, new DDR5 RAM means a new motherboard and CPU if you're still stuck on DDR4, but it was finally getting to the point where you could easily find 16GB or 32GB of some of the best RAM for gaming for under $100 / £100. If you were finally hoping to make the swap to DDR5 this year, I'm so sorry to tell you this, but you'll likely kick yourself for not doing it earlier. Even with Black Friday gaming PC deals beginning to trickle in, RAM, both DDR5 and DDR4, is more expensive than it's been in years. Even our go-to affordable picks are seeing massively inflated costs right now. Crucial Pro DDR5, which you could easily find 32GB of for around $80-$100 a few months ago, is now sitting at $144.99 at Amazon. T-Force's Delta RGB DDR5, which is easily the best budget RAM for most PC builders, is unrecognisable. Again, it's been relatively easy to buy 32GB of this memory for ~$100, but it's currently priced at $150 at Amazon. Looking all the way up and down the price spectrum from the smaller brands to the biggest, RAM is seeing inflated price tags right now, but why am I blaming AI for that, you might ask? Well, allow Chen Libai, the chairman of Adata, to clue you in. He was recently quoted saying, "Our competitors in the fight for supply are no longer our peers, but giant CSPs (cloud service providers)". There's been an enormous increase in AI hardware being produced in recent years, which is more RAM-dependent. RAM works as a speedway for data to be processed by CPUs and NPUs, and since AI uses a lot of data to swiftly produce answers and content, RAM is more vital to these machines. Not only that, but AI data centers are extremely hungry for this RAM as AI models train themselves using vast amounts of data. According to our friends at Tom's Hardware, this could spark a "price apocalypse that could last a decade". Training for these models is done through masses of data, which is dependent on memory and storage - with even hundreds of gigabytes of DRAM being used at a time for one training cluster, according to reports. It doesn't help that, if anything, RAM has been under-produced in the last few years due to a low demand from PC gamers who have been reluctant to upgrade to the latest standard. So when the AI surge began, the manufacturers of DRAM tech were already on the back foot in terms of supply and manufacturing speed. Now, OpenAI has recently signed onto an agreement with Samsung and SK Hynix for up to 900,000 DRAM wafers per month, which, according to Tom's Hardware, is roughly 40% of the global DRAM output of previous years. Seasoned PC gamers might be thinking back to the 2020 days of mining cryptocurrencies, and how that caused a global GPU shortage, which even modern graphics cards are still impacted by in terms of prices. In short, AI is already causing massive surges in DDR4 and DDR5 RAM costs, which is seriously going to hurt consumers as we head into the busiest shopping period of the year. I'll be keeping an eye out for as many low prices as I can, but if you're set on building a new PC at the moment, you may have no choice but to stomach inflated costs, or look at buying second-hand if you can.