"When Qualcomm Does It, We Follow"... Samsung Semiconductor's Innovation Dead [Samsung Crisis Diagnosis]③

Interview with Former Exynos Practitioner
"Must Become an 'AI First Mover' for Prosperity in the Next 50 Years... Hope to Become an 'Open Organization'"

"Does Qualcomm Snapdragon also have that technology? Qualcomm doesn't do it, so why should we (Samsung Electronics)?"

A, a development leader from a partner company who worked for 10 years in the mobile application processor (AP) 'Exynos' development team, the brain of Samsung Electronics' Galaxy phones, and then left Samsung, evaluated Samsung Semiconductor as "innovation is dead." After moving to a partner company, A clearly felt Samsung's problems of being stingy in recognizing the value of external technology and hesitating to make bold investments and collaborations. He sincerely advised that for Samsung Electronics to become strong again, it must break away from conservatism and the culture of departmental silos.

Samsung Electronics Hwaseong Campus view. [Photo by Samsung Electronics]

A, who requested anonymity for the interview, said Samsung Semiconductor ▲ undervalues external technology and ▲ is passive in collaborating with external companies and former Samsung talents. He pointed out, "When talents leave Samsung and start partner companies, Samsung should induce their technologies and products to naturally integrate into Samsung's supply chain through win-win management, but such cases are rare."

As a specific example, he mentioned the case where Puriosa AI, a fabless (semiconductor design specialist) company specializing in neural processing units (NPU) for artificial intelligence (AI), attracted investment from KT instead of Samsung. Puriosa AI's founder, CEO Baek Junho, is a former member of Samsung Semiconductor's (DS) Memory Business Division design team.

A said, "In the past, Samsung handled everything internally even in the fabless field, making it difficult for other companies to do business, but now the situation is different," adding, "Talents who left Samsung have settled in various fields. However, Samsung is passive not only in joint development, investment, and acquisition with external companies but also in collaboration with former Samsung talents."

Based on his experience, A cited Samsung Semiconductor's organizational culture as largely characterized by 'conservatism fearing failure' and 'departmental silo phenomenon.' He said that even if employees propose brilliant ideas, it rarely leads to executive decisions, and "there are almost no cases where people remove their ranks and freely discuss to apply the best ideas."

Application processor (AP) 'Exynos 2400' installed in Samsung Electronics Galaxy S24. [Photo by Samsung Electronics]

Interdepartmental collaboration is also not easy. He shared an experience during his time at Samsung when Samsung Venture Investment's plan to invest in a U.S. startup was canceled due to opposition from the planning team. At that time, the head of the business division (president level) in the DS division proposed, "Let's actively consider investment and acquisition," but it was canceled due to the planning team's objection. As such incidents repeated, business division executives increasingly became reluctant to make challenging investments.

On the other hand, he said that competitors, big tech companies, are overly conscious. During the transition of the global mobile AP market from 32-bit to 64-bit in 2015, Samsung wavered between developing 32-bit and 64-bit products. Initially, they were developing 32-bit products but switched to 64-bit after hearing that Qualcomm was making 64-bit products. A said, "Samsung has closely watched competitors' trends and followed them, adopting a 'fast follower' strategy that achieves high efficiency with fewer personnel," and evaluated this as a result of an organizational culture that does not tolerate failure.

He advised that for Samsung Semiconductor to survive the next 50 years, it must become an 'AI first mover (pioneer).' To do so, he emphasized that conservatism and the culture of departmental silos must be broken. He said, "If Samsung continues to stick to the fast follower strategy like now, the vicious cycle that led to failure in developing high bandwidth memory (HBM) could repeat," and advised, "To become a leader in the AI semiconductor field, Samsung must transform into an 'open organization' that carries out mid- to long-term strategies."