Samsung’s biggest reshuffle boosts AI, robotics integration

Samsung Electronics has promoted 161 staff to executive roles in its largest leadership reshuffle in recent years, the company said on November 25.

The move, which exceeds last year’s 143 promotions, highlights a focus on areas such as AI, robotics, and semiconductors.

The new executives include 51 vice presidents, 93 executive directors, and 16 master engineers.

Notable appointments include Lee Yoon-soo as head of data intelligence at Samsung Research, Lee Sung-jin as vice president for generative AI projects, and Choi Ko-eun as executive director for robotics software.

In its semiconductor division, Hong Hee-il was promoted for work on DRAM and high-bandwidth memory technology.

Samsung also elevated Jacob Zhu, the only foreign national in this round, to lead China sales for its Device Solutions division.

The company said more vice presidents are now in their 40s, and several employees in their 30s have entered senior roles.

🔗 Source: The Korea Herald

🧠 Food for thought

Implications, context, and why it matters.

Samsung promoted Hong Hee-il for work on High Bandwidth Memory (HBM). Reports say its 12‑high HBM3e version with 12 stacked DRAM dies for more capacity and bandwidth passed some NVIDIA tests after an 18‑month delay, with formal approval still pending 1.
A second round of Wafer Acceptance Tests (WATs) on that 12‑high HBM3e for NVIDIA finished 1. WATs are wafer‑level screens viewed as mock runs, not final certification 1.
SK hynix finished HBM4 development and began shipping in Q4 2025, media say NVIDIA is likely a client, while Samsung targets early‑2026 validation (customer qualification) 2.
Yield problems at Samsung, yield being the share of chips that pass testing, tightened dynamic random‑access memory (DRAM) supply and lifted prices 3. SK hynix and Micron push through‑silicon via (TSV) and power‑delivery tweaks as differentiators, though every vendor faces yield hurdles 3.
Samsung promoted Lee Sung-jin to vice president for generative AI, and named Lee Yoon-soo head of data intelligence.
It built Gauss and Gauss2 large language models that support many languages, with three sizes for device capabilities 4.
A LinkedIn summary of its AI roadmap says over 400 million Galaxy devices will get AI features in 2025, with AI in 90% of products by 2030 5. That covers smartphones, tablets, and PCs 5.
Independent software vendors (ISVs) and enterprise software vendors should watch for software development kits (SDKs) or partner programs for Gauss, since mass device reach can unlock business‑to‑business (B2B) deals.