• Samsung has opened the AGI Computing Lab in Silicon Valley, marking its entry into the AI semiconductor logic chip domain, as reported by South Korean media.
  • Led by Senior Vice President Dong Hyuk Woo, formerly of Google, the aim is to create an affordable general AI platform.
  • This move signals Samsung’s shift towards non-memory-based general AI computing chip development, reflecting its strategic focus on core areas of the AI market.

According to South Korean media, Samsung has established the AGI Computing Lab in Silicon Valley, venturing into the AI semiconductor logic chip field.

Dong Hyuk Woo, a former Google developer

Samsung Electronics is overseen by Senior Vice President Dong Hyuk Woo, a former Google developer who joined Samsung in November last year with the aim of constructing an affordable general AI platform. During his tenure at Google, Dong Hyuk Woo was involved in launching the Edge TPU project, promoting the commercialisation of various machine learning applications on Tensor SoC for device-side use cases; he was also one of the first three engineers to contribute to building Google’s first cloud TPU with training capabilities.

Also read: OpenAI’s Altman to meet chipmakers Samsung

Samsung targets the core areas of the AI market

Dong Hyuk Woo’s final project at Google was the Gemini Nano AI model. Samsung is expanding the organisational structure of the AGI Computing Lab, providing positions for key personnel such as micro-architects. South Korean media suggests that up to now, Samsung Electronics’ traditional role in the AI semiconductor market has been as a supplier of high-performance storage chips (such as HBM) for auxiliary computing, rather than as a developer of logic chips; now, Samsung is fully engaged in the development of non-memory-based general AI computing chips, demonstrating its determination to target the core areas of the AI market.

Professor Jung Woo-sung from Pohang University of Science and Technology in South Korea remarked, ‘Under the leadership of semiconductor giants, a new landscape is emerging in the field of general AI chips. Market changes will unfold unpredictably due to existing products’ inability to meet demands.”