en.Wedoany.com Reported - Samsung Electronics' network business recorded revenue of $407.5 million (approximately 600 billion won) in the first quarter of 2026, a significant decline from approximately $543 million (around 800 billion won) in the same period of 2025. This marks one of the weakest first-quarter financial performances for the business since 2018. Hampered by the broader environment of persistently tightening capital expenditure among global telecom operators, the head of Samsung's network business acknowledged the industry is facing "headwinds" and has already begun implementing "structural cost controls" to secure profitability.

Facing the downturn in the telecom market, Samsung is rapidly shifting its strategic focus to the AI-driven data center optical communication market. During its first-quarter earnings call on April 30, 2026, Samsung clearly stated that the urgent demand for high-bandwidth, low-power data transmission in AI data centers has become the core engine for its foundry business growth. Suk Chae Kang, Executive Vice President of Samsung's foundry business, revealed during the call: "We are actively driving a structural transformation, expanding our application portfolio from mobile to AI HPC, automotive, and aerospace sectors." He further confirmed that the foundry division has successfully secured an order from a major optical communication module manufacturer and plans to begin mass production for them in the second half of 2026. Boosted by this news, Samsung's stock price showed an upward trend following the earnings release.

Samsung's determination to enter the optical communication market was already evident when it officially announced its entry into the silicon photonics foundry business in March 2026. At the Optical Fiber Communication Conference and Exhibition (OFC 2026) held in Los Angeles in March 2026, Samsung unveiled its silicon photonics commercialization roadmap for the first time. According to the plan, the company has completed the development of its Process Design Kit (PDK) and possesses the immediate capability to accept customer Photonic Integrated Circuit (PIC) designs on its 300mm wafer platform. Its technology roadmap indicates plans to launch an optical engine based on Thermo-Compression (TC) bonding in 2027 and to provide a complete Co-Packaged Optics (CPO) solution by 2029, integrating optical modules directly into AI chip packaging. Suk Chae Kang added during the earnings call: "We are not only developing silicon photonics components but also advancing full-stack technologies including co-packaged optics."

Compared to the silicon photonics alliance of industry leaders like U.S.-based Nvidia and TSMC, Samsung emphasizes its unique "one-stop" delivery capability, a differentiated service integrating CPO, logic processes, and High Bandwidth Memory (HBM). Jaejune Kim, Executive Vice President of Samsung's memory business, stated on the same conference call that HBM sales are expected to more than triple year-over-year in 2026. Suk Chae Kang also mentioned that regarding the advanced 2-nanometer process, the company is in cooperation discussions with multiple AI/HPC customers and anticipates achieving concrete results in the near future.

Despite the promising prospects in the AI sector, Samsung's traditional telecom equipment sales within its network business are experiencing a harsh winter. Even though the global Radio Access Network (RAN) market stabilized after 2025, Samsung still fell into its lowest performance trough in nearly eight years this quarter. The earnings statement did not detail a specific revival blueprint for its network business, only indicating a commitment to "overseas sales expansion" to address the bottleneck of limited domestic market size. Clearly, Samsung is attempting to use its emerging optical and AI business curves to hedge against the cyclical downturn of traditional communication equipment, aiming to secure a more advantageous position in the next wave of competition for 6G and AI computing power infrastructure.

This article is compiled by Wedoany. All AI citations must indicate the source as "Wedoany". If there is any infringement or other issues, please notify us promptly, and we will modify or delete it accordingly. Email: news@wedoany.com