According to a recent report by the South Korean media Digital Daily, Apple proactively approached Changxin Storage to negotiate DRAM procurement, hoping to secure lower prices than Samsung and SK Hynix, but was directly refused by Changxin. Changxin's stance is clear: its quotes will not be lower than those of the Korean companies, and may even be higher.
The reason is simple: Chinese domestic companies like Huawei and Xiaomi have already locked in Changxin Storage's production capacity through long-term contracts, so Changxin Storage does not need to accommodate Apple's stringent conditions.
Five years ago, this would have been unimaginable. With over 200 million iPhones sold annually, plus the massive shipment volumes of Macs and iPads, Apple has long been the most powerful buyer in the global consumer electronics supply chain.
Entering Apple's supply chain was the dream of countless domestic manufacturers. To secure orders, price cuts, building dedicated production lines, and complying with stringent requirements were standard practices.
Companies like Luxshare, Goertek, and Lens Technology grew from small factories into industry leaders by riding the "Apple chain." But today, Changxin won't even make a price concession. This isn't arrogance; it's a change in the game rules and bargaining power.
Why Is Apple So Eager to Approach Changxin? An Alternative Forced by Price Hikes
At its core, Apple humbling itself to contact a Chinese storage manufacturer is a move forced by this wave of memory price increases.
The explosion of AI large models has made HBM a necessity for compute servers, with profits several times that of ordinary DRAM. To earn more, Samsung and SK Hynix are shifting more and more capacity to HBM, the high-bandwidth memory most needed by AI servers. Customers are even willing to sign three-to-five-year long-term agreements to lock in capacity.
Data shows that the proportion of HBM wafers at SK Hynix has surged to 29.2%, Samsung to 23.4%, and Micron to 18.8%. The result is that the supply of traditional DRAM is becoming increasingly tight. Subsequently, the entire memory market has re-entered a price hike cycle.
How much have prices risen for general-purpose DRAM? Counterpoint's data shows: the price of 64GB server memory multiplied by 3.5 times from Q3 2025 to Q1 2026, with a cumulative annual increase of nearly 490%.
The rise in mobile LPDDR5X is even more dramatic, with the highest single-quarter increase reaching 83%. The cost of a single 12GB memory chip is nearly 90% more expensive than a year ago.
The direct consequence of these price hikes is that the proportion of memory in the BOM cost has spiraled out of control. Previously, memory chips accounted for only 10% to 15% of a phone's total cost; now that has soared directly to 30% to 40%, and for some high-memory models, it's approaching 50%.
The price of the top-tier iPhone model with doubled memory directly increased by over three thousand yuan, largely driven by rising memory costs. Even with Apple's thick profit margins, it can't withstand such increases.
Apple's previous playbook was: have three suppliers bid, give orders to the lowest bidder, and even use "Chinese suppliers" as a bargaining chip to lower prices. Now, the three giants are focused on earning money from AI. No matter how large Apple's orders are, they aren't as attractive as NVIDIA's HBM orders. With fewer and fewer alternative options, its bargaining power naturally weakens.
Approaching Changxin was originally Apple's attempt to add another supplier, gain another bargaining chip for price pressure, and diversify supply chain risks. Apple likely expected Chinese manufacturers to proactively lower prices to join the Apple chain, as in the past. But unexpectedly, Changxin refused.
Behind Changxin's Rejection of Apple's Price Pressure
Changxin isn't acting on a whim. First, its capacity has long been fully booked by domestic clients.
Smartphone manufacturers like Huawei, Xiaomi, OPPO, and vivo, and internet giants like Tencent, Alibaba, and ByteDance have already reserved Changxin's mainstream production capacity through long-term agreements. Major companies have signed hundreds of billions in three-to-five-year long-term orders for server DRAM. There's simply no need to compete for Apple's orders by lowering prices.
On another front, Changxin has now completed a full product portfolio from DDR4 to DDR5, and from LPDDR4 to LPDDR5X.
After entering the second half of 2025, the yield rate for Changxin's DDR5 products has broken through 90%. The mass production yield rate for its 17nm process has also stabilized above 90%, validated through large-scale use in Huawei and Xiaomi's flagship phones.
The industry generally considers the passing line for mass production yield to be around 85%. Samsung's yield rate for the same generation process is roughly 92% to 93%. This means the gap between the two sides has narrowed to a very small range.
Previously, domestic memory competed by being 20% cheaper because there was a technological generation gap. Now that performance is nearly equal, why should it still be sold cheaper?
Even the South Korean media admits it. A few years ago, they were saying Chinese memory was impacting the market with low prices. Now, the discussion has reversed to Changxin's quotes being higher than Korean companies—this in itself is a reversal of industry standing.
Behind Changxin's rejection of Apple's price pressure, there might also be a risk calculation.
The glow of the Apple chain is as dazzling as its pitfalls are deep. At its peak, Apple orders accounted for 20% of O-Film's revenue, and its market value approached 70 billion yuan. After being removed from the supply chain in 2021, O-Film accumulated losses of 9.75 billion yuan over three years, with unrealized losses still exceeding one-third of its paid-in capital.
Wingtech invested heavily to secure Apple's MacBook assembly business. Later, after being placed on the Entity List, all that investment was wasted, and it eventually sold the entire assembly line at a discounted price to Luxshare.
With these cases in front of them, capable companies with massive order books will reassess the weight of the Apple chain. For one customer, to lower prices, build dedicated production lines, modify specialized processes, and highly bind capacity—only to potentially face a listing on some list one day, rendering billions in investment worthless.
Changxin is already a semiconductor company of key concern to the US. The political risk of entering Apple's supply chain outweighs the commercial benefits. Rather than lowering prices, expanding capacity, and bearing policy risks for uncertain orders, it's better to steadily hold pricing power, technological iteration, and capacity expansion in its own hands.
In other words, companies used to scramble to join the Apple chain because it meant making big money and growing fast; now, the rewards and risks are completely out of balance.
AI Deprives Apple of Pricing Power, A Reversal from Buyer's to Seller's Market
Changxin's rejection of Apple signifies that the AI wave has stripped Apple of its pricing power.
For the past two decades, the rules of consumer electronics were set by terminal manufacturers. Apple defined the products, suppliers handled manufacturing; Apple set prices, suppliers undercut each other for orders. Whoever controlled the brand and channels controlled the profit distribution in the industrial chain.
Contract manufacturers were lucky to have a 10% gross margin, while Apple took over 80% of the profits—this was the industry norm.
But with the arrival of the AI wave, this rulebook has been directly rewritten. The speed of AI demand explosion far outpaces the speed of capacity expansion.
Now, what's scarce is not terminal brands, but upstream core capacity. A single production line requires an investment of tens of billions and a construction period of at least two to three years.
Previously, clients chose suppliers; now, suppliers choose clients. Previously, contracts locked in prices and capacity; now, they lock in capacity but not prices, with increases following market rates.
Samsung and SK Hynix dare to shift capacity to HBM, unafraid of offending Apple, precisely because AI clients offer higher prices, pay more readily, and sign more stable long-term agreements. Changxin dares not to accommodate Apple because domestic demand is sufficient to support its development, and there's no need to bear additional risks for Apple.
Chinese semiconductor enterprises are transitioning from "low-cost substitutes" to "equal suppliers."
Previously, when we talked about domestic substitution, it implicitly meant being 20% cheaper than foreign counterparts, competing on cost-effectiveness. That's changing now. With similar performance and similar prices, they can even be more expensive during tight supply and demand.
The reason is that Chinese companies can offer more stable supply, and gain pricing power through technology and capacity.
In high-end HBM and the most advanced process nodes, we still have a three-to-four-year generation gap, and the current global market share is only around 8%. There's still a long way to go.
But this event shows that China's memory industry has moved from "can we make it?" to "can we negotiate on equal terms?"; it shows that Chinese semiconductor companies no longer view entering the Apple chain as the ultimate goal.
On the other hand, Apple is actively lobbying the US government to secure procurement qualifications. To some extent, this is also a signal Apple is sending to Changxin. Apple hopes to prove that this cooperation is not short-term exploitation, but a genuine desire to establish a long-term partnership.
The problem, however, is that Apple's most important card in the past—using long-term orders and massive purchase volumes to secure low prices and priority supply—is slowly losing its effectiveness.
The fading of the Apple chain myth is an inevitable outcome of China's industrial upgrading.
When Chinese enterprises possess their own core technology and a stable base of key clients, they no longer need to treat any single large customer as a lifeline. Moving from scrambling for orders on their knees to doing business standing up is an advancement in strength and a shift in discourse power.
The ultimate logic of the industrial chain has never changed: whoever controls scarce core capacity and technology controls pricing power. And for Apple, this is the first time it has lost pricing power when facing a mainland Chinese supplier. The times may truly have changed.
This article is from the WeChat public account "Hot Spot Micro Commentary" (ID: redianweiping), author: Wang Xinxi
