Samsung Electronics is about to pull back the curtain on what could be one of the most jaw-dropping quarters in semiconductor history. The company’s preliminary Q2 2026 earnings guidance, expected around July 7, will give investors their first real look at how the AI chip boom is reshaping the world’s largest memory maker.
Brokerage estimates peg Samsung’s Q2 operating profit at roughly 86 trillion KRW, which would represent a year-over-year increase of approximately 1,740%. That’s not a typo.
The numbers behind the AI chip supercycle
To understand why analysts are this bullish, you need to look at Samsung’s Q1 2026 results. The company posted a consolidated operating profit of 57.2 trillion KRW, roughly $38-39 billion, marking a 756% increase compared to the same period a year earlier.
The engine behind those numbers was Samsung’s Device Solutions division, the unit responsible for memory chips, system LSI, and foundry operations. DS revenue hit 81.7 trillion KRW in Q1, an 86% jump quarter-over-quarter. The semiconductor unit contributed over 93% of Samsung’s overall profits for the quarter.
The memory business, specifically DRAM, NAND, and high-bandwidth memory (HBM), achieved record sales driven by hyperscaler spending on AI infrastructure. For Q2, projected revenue is approaching 170 trillion KRW, supported by average selling price increases of 40-60% quarter-over-quarter for memory products.
What investors should be watching
The July 7 preliminary guidance will likely confirm whether the momentum Samsung built in Q1 is accelerating or plateauing. HBM, the specialized memory used in AI accelerators like Nvidia’s GPUs, has been the highest-margin product in Samsung’s portfolio. How much of the revenue growth is coming from HBM versus conventional DRAM and NAND will tell investors whether Samsung is capturing its fair share of the AI premium or ceding ground to competitors like SK Hynix, which has been widely regarded as the HBM market leader.
A 40-60% quarter-over-quarter increase in memory pricing is extraordinary. If Samsung’s guidance suggests these price increases are sustainable into the back half of 2026, it would imply that the supply shortage is structural rather than cyclical.
Samsung has warned that supply-demand deficits in the memory chip market could actually worsen heading into 2027 as AI adoption accelerates globally. If the world’s largest memory maker is saying it can’t produce enough chips to meet demand, that has cascading effects across every industry that depends on computing power.
Samsung’s foundry business, which manufactures chips for other companies, has been struggling to keep pace with TSMC. Any signs of improvement in foundry win rates or yield performance could shift the semiconductor supply picture in meaningful ways.
Given that the consensus estimate already bakes in a nearly 18x year-over-year profit increase, the bar for a positive surprise is genuinely high. A miss, even a modest one, could trigger a reassessment of how sustainable the AI chip boom really is.