Finance
Samsung set for highest Q3 profit in three years as AI demand lifts chip prices
Published by Global Banking & Finance Review®
Posted on October 12, 2025
3 min readLast updated: January 21, 2026
Global Banking & Finance Review®
Company
Published by Global Banking & Finance Review®
Posted on October 12, 2025
3 min readLast updated: January 21, 2026
Samsung anticipates its highest Q3 profit in three years, driven by AI demand boosting memory chip prices. Challenges remain with HBM chip supply delays.
By Heekyong Yang
SEOUL (Reuters) -Samsung Electronics is expected to post its highest third-quarter profit since 2022, driven by higher memory chip prices supported by server demand as customers rebuild inventories, analysts' estimates showed.
The world's biggest maker of memory chips is projected to report an operating profit of 10.1 trillion won ($7.11 billion) for the July-September period, according to LSEG SmartEstimate from 31 analysts, which is weighted toward those who are more consistently accurate. This would be up 10% from a year earlier.
Analysts attributed the recovery mainly to better conventional memory chip pricing, which would offset weaker sales volumes of high-bandwidth memory (HBM) chips as Samsung has yet to supply its latest HBM products to Nvidia.
HBM chips, critical for artificial intelligence (AI) development, are designed to reduce power consumption and process large datasets by stacking chips vertically.
Analysts said demand for memory chips, particularly from hyperscalers and AI-related investments for services such as ChatGPT, have put more workload on general servers, thus boosting conventional memory chip prices.
Prices of some DRAM chips, widely used in servers, smartphones and PCs, jumped 171.8% in the third quarter from a year earlier, according to TrendForce data.
While Samsung's conventional memory business performed well, analysts said delays in supplying its latest 12-layer HBM3E chips to Nvidia have hurt its profit and share price.
Rivals SK Hynix and Micron have gained more from AI-driven demand, while Samsung's exposure to China, where advanced chip sales are restricted by the United States, has constrained its growth.
Analysts said market sentiment toward Samsung's shares and chip business, including both memory and contract chip manufacturing, is expected to improve as it secures supply deals with major customers such as OpenAI and Tesla.
Samsung shares have risen more than 43% following its announcement of a chip supply deal with Tesla in July.
During OpenAI CEO Sam Altman's visit to South Korea earlier this month, Samsung, SK Hynix and OpenAI announced partnerships to supply advanced memory chips to the Stargate project.
The AI chip deal between OpenAI and AMD, one of Samsung's major HBM customers, would also benefit Samsung, said Ryu Young-ho, a senior analyst at NH Investment & Securities.
Ryu added that Samsung's $16.5 billion foundry deal with Tesla has lifted expectations that Samsung's struggling contract chip manufacturing business could win more orders from major tech firms if the company delivers the project as planned.
While recent AI-driven supply deals signal a positive outlook for Samsung, analysts cautioned that uncertainties remain, including potential U.S. tariffs on chips and China's tightened export controls on rare earth materials used in advanced chips and manufacturing equipment.
In September, Micron said it expects to sell out all of its HBM chips for calendar year 2026 in the coming months due to strong demand.
Samsung will announce its estimates on revenue and operating profit on Tuesday, with full results due later this month.
($1 = 1,420.1000 won)
(Reporting by Heekyong Yang; Editing by Jacqueline Wong)
Operating profit is the income a company generates from its core business operations, excluding deductions of interest and taxes. It reflects the efficiency of a company's operations.
Memory chips are electronic components that store data for computers and other devices. They are crucial for the performance of devices, especially in data processing and storage.
Artificial intelligence (AI) refers to the simulation of human intelligence in machines programmed to think and learn. AI technologies are used in various applications, including data processing and automation.
Hyperscalers are large-scale cloud service providers that operate massive data centers and provide scalable computing resources. They are key players in the technology and data storage sectors.
High-bandwidth memory (HBM) is a type of memory used in computing that provides high data transfer rates and is designed for applications requiring large amounts of data processing, like AI.
Explore more articles in the Finance category
