Memory Chips Enter Supercycle, Citi Sets $840 Price Target for Micron. Will Upcoming Nvidia Earnings Confirm Chip Bull Market?

The current wave of price hikes in the DRAM market is spreading rapidly. Samsung took the lead by raising product prices by 100% in the first quarter of 2026, and Micron plans to follow suit in the second quarter of the same year with an increase of approximately 40%.

The core driver of this price rally is the supply-demand imbalance in standard commercial DRAM (non-HBM). According to the silicon systems sales outlook from equipment manufacturer Applied Materials (AMAT), DRAM bit supply growth is expected to reach only about 30% by the end of 2026. Such an increment is clearly insufficient to match the explosive demand growth from AI data centers in 2027; consequently, the industry will need to invest in new wafer capacity to fill the gap.

Forecasts from Citi corroborate this trend, indicating that average DRAM prices will surge by approximately 200% year-over-year in 2026, while NAND prices are expected to rise by about 186%. This suggests the memory industry is currently in the midst of a historic strong pricing cycle.

Meanwhile, the supply tightness in HBM persists, and expectations for price hikes in 2027 continue to mount. This is one of the primary reasons for the upward revision of Micron's valuation.

Producing HBM requires three to four times more wafer capacity than standard DRAM. However, the current profitability gap between HBM and commercial memory is not yet sufficient to entice memory manufacturers to convert capacity on a large scale or add new production lines. As a result, manufacturers remain generally cautious regarding supply, which keeps HBM in a state of persistent shortage.

Amid constrained supply, HBM prices have further upside potential in 2027. Nevertheless, memory manufacturers are expected to maintain disciplined supply management to prevent excessive prices from causing AI data centers to reduce their HBM procurement volumes.

In the global memory chip manufacturing sector, South Korea's SK Hynix and Samsung Electronics hold leading positions. SK Hynix is the primary supplier of HBM chips for Nvidia, while Samsung and Micron are competing directly with it in this space.

Citi analysts noted that the HBM supply crunch is ongoing. Due to the 3-to-4-times wafer conversion ratio between HBM and standard DRAM, coupled with the margin differential between the two, memory manufacturers currently lack the incentive to convert or add HBM capacity. This 3-to-4 wafer conversion ratio typically refers to the comparison of physical switching costs or processing difficulty between 3-inch and 4-inch semiconductor substrates.

Analysts also mentioned that, given the current tightness in HBM capacity, prices are likely to continue rising in 2027. Memory manufacturers are expected to exercise restraint when increasing supply to prevent AI data centers from reducing HBM content next year.

However, it is worth noting that risks on the demand side cannot be overlooked.

Cisco ( CSCO) has already slashed DRAM usage by 50% across more than 20 product lines, including wireless products, due to the pressure of rising memory prices. This case serves as a warning to the industry that excessive memory prices could dampen end-market demand; memory manufacturers must find a balance between the pace of price hikes and the maintenance of demand.

The accelerating penetration of AI applications is reshaping the landscape of the memory chip industry, a trend clearly evidenced in Micron's fiscal Q1 2026 earnings call. Management emphasized during the call that demand for high-capacity data center SSDs for AI infrastructure, KV cache deployments, and PCIe Gen6 SSDs compatible with NVIDIA's ( NVDA) AI computing clusters is experiencing explosive growth, signaling that the demand scope for AI-related storage chips has far exceeded the previous expectations of many Wall Street analysts.

In addition to Citi, Melius Research analyst Ben Reitzes also significantly raised Micron's price target from $700 to $1,100, a 57% increase. Based on this calculation, Micron's current stock price still has approximately 65% upside potential, while Reitzes maintained a Buy rating on the stock.

He stated that his research team is fully confident in the memory chip sector and the AI semiconductor industry as a whole. They not only raised Micron's valuation but also simultaneously increased the long-term valuations of several chip companies including AMD ( AMD ), Intel ( INTC ), Marvell ( MRVL ), Qualcomm ( QCOM ), SanDisk ( SNDK ), and others, believing that chipmakers will continue to capture value share from traditional software companies.

From the perspective of long-term storage market dynamics, JPMorgan ( JPM) analyst Jay Kwon expects the upward trend in memory chip prices to persist at least until the end of 2027. The existence of long-term supply contracts is expected to bring more stability to this historically highly cyclical industry.

At present, NVIDIA's upcoming earnings report is viewed by the market as the core catalyst for the next leg of the rally.

Currently, investors are diversifying capital from the core NVIDIA GPU trade into previously overlooked AI beneficiary sectors, whose supporting roles in AI infrastructure have not been fully recognized by the market. Investors are generally betting that as the AI industry transitions from the training phase to the agentic application phase, demand for other AI components such as memory chips, alongside computing chips, will also experience explosive growth.

NVIDIA's fiscal first-quarter 2027 earnings report (covering results through April 26), scheduled for release after the U.S. market close on May 20, will serve as a core variable dictating the subsequent direction of U.S. equities, as it will provide the most direct validation of whether global demand for AI infrastructure remains robust. Leading Wall Street institutions generally expect NVIDIA to once again deliver revenue and profit growth that significantly outperforms market expectations.

This content was translated using AI and reviewed for clarity. It is for informational purposes only.