Samsung Electronics said on Tuesday it wants to undertake mergers and acquisitions – M&As – and will concentrate as much on design as on manufacturing in its bid to become the dominant global player in the semiconductor business.
Samsung revealed in April that it aims to be the world’s top corporate in semiconductors – known as logic chips – by 2030.
To this end, the group plans to invest 60 trillion won ($50.5 billion) in facilities and 73 trillion won ($61.5 billion) on research and development (R&D) over the next decade.
Third global No 1 spot
If successful in grabbing that top spot, the Korean electronics giant would add a third “number one” title to its portfolio, as it is already the world’s largest smartphone maker and the world’s biggest manufacturer of memory chips.
However, memory chips are not as high value-added products as logic chips, and, being commodity products, are vulnerable to cyclic sector downturns.
In a press conference on Tuesday, Kang In-yup, the head of Samsung’s System LSI (Large-scale Integration) Business, elaborated on how Samsung will reach its goal.
Samsung is “open to M&As,” Kang said. “It is difficult to be the top on our own, as product portfolios in the logic chip sector are diverse. [So] we are open to big or small M&As to get the missing technologies or talent that we don’t have.”
He also gave some detail on where Samsung’s main efforts will be.
“We are going to take the top spot in the fabless business,” Kang said. That indicates that Samsung may not be the top player in all areas of the sector – most notably chip production, Samsung’s core competency – but its goal is to be top in the design and development of system semiconductors.
Focusing on the brain
The Korean electronics giant also announced its plan to strengthen its position in Neural Processing Units (NPUs). To that end, Samsung Electronics will expand its research workforce in that field by more than 10 times – to more than 2,000 personnel – by 2030.
An NPU is a key “brain” and the essential logic chip for AI, as it is optimized for deep learning algorithm computation. Samsung introduced its first NPU when the firm developed its mobile processor, Exynos 9820, last year.
It plans to expand the deployment of NPU into all IT fields, ranging from flagship SoC (system-on-chip) products for mobile devices, automotive electronic components such as infotainment systems and advanced driver assistance systems to deep-learning NPUs that enhance data center’s performance in processing big data and IoT.
Samsung Electronics said it would nurture the NPU business with its own technologies, but could also take over related businesses via mergers and acquisitions.