On the first trading day of July, the crypto market continued its weak consolidation trend. According to Gate market data, Bitcoin broke below the psychological threshold of $60,000, settling at $58,554.7. Ethereum also slipped below the $1,600 mark, closing at $1,574.07. The total crypto market cap held steady around $2 trillion, with Bitcoin’s market dominance exceeding 57%. Over the past 24 hours, liquidations across the market reached approximately $249 million, putting significant pressure on bullish positions. The Fear & Greed Index dropped to the 11-16 range, remaining in a state of extreme fear.
In stark contrast to the chill in the crypto market, the semiconductor memory sector is experiencing a wave of aggressive upward revisions in expectations. On June 30, Goldman Sachs released its DRAM sentiment indicator, raising its forecast for Samsung Electronics’ HBM price growth in 2027 from 14% to a striking 44%. This adjustment is substantial—an increase of 30 percentage points. Notably, Goldman also stated that even this revised forecast "may still not fully reflect the upside risk," suggesting that the 44% estimate "might not be the end point."
Where does this 44% come from? Why is Goldman making such a dramatic upward revision at this particular moment? To answer these questions, we need to break down the three core logic chains in Goldman’s report.
Logic One: Traditional DRAM Spot Price Surge Is Redefining the HBM Pricing Anchor
The most immediate trigger for Goldman’s upward revision comes from the strong performance of traditional DRAM spot prices.
According to Goldman’s tracking, as of June 26, DDR5 spot prices have risen 20% since May 1, with a 25% premium over May contract prices. DDR4 spot prices increased 11% in the same period, commanding a staggering 45% premium over contract prices. The latest quote on July 1 shows the average price of DDR5 16Gb holding steady at $46.833, up 0.21% from the previous trading day, with spot prices still running high.
Spot prices significantly outpacing contract prices are typically seen in the memory industry as a leading indicator for subsequent contract price increases. Goldman’s report clearly states: "The current strong performance of traditional DRAM spot prices will inevitably become a key reference when discussing next year’s HBM pricing."
This logic chain is straightforward. HBM contract negotiations don’t happen in isolation—when memory manufacturers and cloud providers or AI chip companies discuss next year’s HBM supply contracts, traditional DRAM market prices serve as a crucial benchmark. As DDR5 and DDR4 prices continue to strengthen, memory manufacturers gain a firmer bargaining position. HBM, being a higher-end and more scarce product, may see its premium over regular DRAM expand rather than shrink.
Goldman further projects in its model: If HBM supply remains tight and the price gap between traditional DRAM and HBM continues to widen, there’s potential for the 44% forecast to be revised even higher. This is a self-reinforcing logic—the more traditional DRAM prices rise, the greater the room for HBM price hikes.
Logic Two: Cross-Validation from Industry Chain Data—Price Increases Are Not an Isolated Phenomenon
If price expectations are based solely on theoretical modeling, their persuasive power is limited. Another layer of support in Goldman’s report comes from robust May export and revenue data.
Official Korean data shows that in May 2026, Korea’s total exports reached $87.8 billion, up 53.2% year-over-year, setting a new monthly record. DRAM exports totaled $18.6 billion, soaring 369.8% year-over-year and up 21% month-over-month, also hitting a historical high. NAND exports rose 206.8% year-over-year. Goldman notes that this performance is driven by rising memory prices combined with strong demand from major tech companies in the US and China.
Taiwan’s supply chain data is even more direct. Nanya Technology’s consolidated revenue for May 2026 was NT$27.67 billion, up 8.55% month-over-month and a staggering 730.14% year-over-year, marking ten consecutive months of triple-digit annual growth. Goldman’s tracked distributor, Supreme Electronics, saw May revenue jump 253% year-over-year and 54% month-over-month.
The server supply chain is also booming. Taiwanese server ODMs (Inventec, Quanta, Wiwynn, Wistron) collectively posted a 53% year-over-year increase in May revenue, driven by rapid growth in rack-level AI server and ASIC AI server shipments. Global server BMC supplier Aspeed saw May revenue rise 69% year-over-year—remarkably, this was achieved on top of a 75% high base from the previous year.
This data points to one conclusion: AI server and traditional server demand is absorbing higher memory prices. Price increases are no longer just numbers on a quote screen—they’re reflected in manufacturers’ revenues and export figures. As long as downstream buyers are willing to accept higher prices, memory manufacturers will have strong negotiating power for 2027 HBM contracts.
Logic Three: Rigid Supply + Resilient Demand—The Supply-Demand Gap Won’t Close Quickly
The third pillar supporting HBM price hike expectations comes from structural constraints on the supply side.
Currently, over 95% of global DRAM capacity is controlled by Samsung, SK Hynix, and Micron. Since HBM gross margins exceed 70%—far higher than traditional DRAM’s roughly 40%—these three giants continue to prioritize limited 12-inch wafer capacity for HBM production, which in turn squeezes supply for DDR4, DDR5, and other general-purpose memory.
HBM’s supply elasticity is much lower than regular DRAM. Capacity conversion is more difficult, customer certification cycles are longer, and building additional production capacity requires massive investment and often takes years. This means that even with clear price signals, supply-side response is naturally constrained.
Demand shows no signs of weakening. Micron’s entire 2026 HBM supply has already been sold out under fixed-price contracts. TrendForce estimates that SK Hynix will maintain about 50% HBM market share in 2026, Samsung around 28%, and Micron about 22%. Although the three giants have aggressive expansion plans—Samsung aims to boost HBM capacity by 50% in 2026, targeting 250,000 wafers per month—the supply gap is unlikely to be fully closed before 2027, given the continued expansion of AI server demand.
Goldman estimates that Samsung’s average DRAM selling price in Q2 2026 will rise roughly 46% quarter-over-quarter, then slow to 15% and 7% in Q3 and Q4, respectively. Prices may continue to climb, but the sharpest increases are likely in Q2. Goldman maintains a "Buy" rating on Samsung Electronics, with a 12-month target price of KRW 480,000 for common shares and KRW 360,000 for preferred shares.
Risk Boundaries: 44% Is a Forecast, Not a Commitment
No analysis is complete without addressing the risks. Goldman’s report includes clear risk warnings.
First, it’s important to clarify that 44% is Goldman’s forecast for Samsung’s HBM price growth in 2027—it’s not an actual quoted price, nor is it Samsung’s official guidance. The forecast assumes continued tight HBM supply and strong traditional DRAM prices. If these assumptions change, the forecast will be revised.
Goldman trader Ippei Yamauchi, in another report, cautions that the recent tech stock sell-off reflects not just a short-term correction, but structural challenges facing AI infrastructure and the memory supply chain. The report highlights three major risks:
First, HBM supply and demand may shift in 2027-2028. As Samsung, Micron, and SK Hynix aggressively expand capacity, HBM supply will increase rapidly, and the premium from supply shortages is expected to peak during this period.
Second, Chinese memory players are rising quickly. According to the Financial Times, Apple is actively lobbying the US government to approve memory purchases from China’s CXMT. If future policy permits, this could directly impact the market position of existing DRAM suppliers.
Third, the AI investment boom is showing signs of cooling. Qualcomm is exploring ways to reduce reliance on HBM, and Nvidia continues to optimize memory efficiency. OpenAI recently lowered service prices, raising fresh doubts about when massive AI investments will translate into real profits.
Additionally, smartphone demand remains a drag. While China’s smartphone shipments in May rose 19% year-over-year, cumulative shipments in 2026 are roughly flat compared to 2025, with Q2 shipments expected to drop 14% year-over-year. Price increases themselves may suppress some consumer electronics demand—AI servers and cloud providers can tolerate higher memory prices, but if phones and PCs can’t pass on costs smoothly, they may cut purchases or delay restocking.
Two Markets, One Question
Let’s shift focus back to the crypto market. Bitcoin fell below $60,000 on July 1, with a nearly 20% cumulative decline in Q2. In June, US spot Bitcoin ETFs saw net outflows of about $4.06 billion—the largest monthly redemption since launch.
Rising memory chip prices and falling crypto asset prices may seem like two separate worlds, but they’re both grappling with the same question: How are asset prices being redefined under the dual constraints of liquidity and end-user demand? AI computing power needs underpin the logic behind HBM and DRAM price hikes, but weak consumer electronics draw boundaries around that logic. The crypto market faces similar tension—institutional funds are steadily flowing out at the ETF level, and on-chain data shows some long-term holders are engaging in "capitulation selling."
Investors in both markets are waiting for the same answer: Can end-user demand support current valuations?
Conclusion
Goldman raised its forecast for Samsung’s HBM price growth in 2027 from 14% to 44%, based on three interlocking logic chains: traditional DRAM spot price increases are redefining HBM’s pricing reference; Korean export and Taiwanese manufacturer revenue data cross-validate the bullish outlook; and rigid supply combined with resilient demand means the supply gap won’t close quickly. Goldman also clearly notes that the 44% forecast "might not be the end point."
However, 44% remains a forecast, not an actual price. Whether this prediction materializes depends on the sustainability of AI server demand, the pace of HBM supply expansion, and whether consumer electronics can absorb higher memory prices. Goldman itself highlights three structural risks: shifting HBM supply-demand dynamics, the rise of Chinese manufacturers, and cooling AI investment.
For crypto market investors, the pricing logic of memory chips offers a window into the health of the AI industry chain. As crypto assets remain in a weak consolidation phase, understanding the cost structure and pricing dynamics of computing infrastructure may be more valuable than simply watching price charts. The logic behind price increases is clear, but whether it plays out as expected will ultimately be decided by the market.
FAQ
Q1: Why did Goldman raise its forecast for Samsung HBM prices from 14% to 44%?
The primary reason is the sustained strength in traditional DRAM spot prices. DDR5 spot prices have risen 20% since May, with a 25% premium over contract prices; DDR4 is up 11%, with a 45% premium. Spot prices typically lead contract markets, and the current strong performance of traditional DRAM will be a key reference in next year’s HBM pricing negotiations.
Q2: Is the 44% price increase forecast an actual price hike?
No. The 44% figure is Goldman’s forecast for Samsung’s HBM price growth in 2027—it’s not a realized quote or official company guidance. Goldman also believes that if HBM supply remains tight and the price gap with traditional DRAM widens further, this forecast could be revised higher.
Q3: Besides traditional DRAM price increases, what other data supports this forecast?
Korea’s DRAM exports in May rose 369.8% year-over-year and 21% month-over-month, setting new records. Nanya Technology’s May revenue jumped 730% year-over-year, marking ten consecutive months of triple-digit growth. Taiwanese server ODMs saw collective revenue rise 53% year-over-year, and server BMC supplier Aspeed posted a 69% year-over-year revenue increase. Industry chain data is cross-validating the price hike logic.
Q4: Did Goldman highlight any risks?
Goldman’s trader pointed out three structural risks in another report: HBM supply-demand may shift in 2027-2028; Chinese memory manufacturers (such as CXMT) are rising; and the AI investment boom is showing signs of cooling. Additionally, weak smartphone demand and price hikes themselves may suppress consumer electronics demand.
Q5: What does the HBM price hike mean for the crypto market?
HBM is the core memory paired with GPUs in AI servers, and its price trend reflects changes in the cost of AI computing infrastructure. In the crypto market, AI-related projects and mining hardware costs are indirectly linked to chip prices. Furthermore, the health of the memory chip sector and global risk asset pricing are connected via liquidity, making it a useful indicator for gauging sentiment in the tech industry chain.
