Review of Attending SK Tech AI 2024
SK AI Summit 2024, the largest conference in South Korea that brings together SK Group's AI capabilities, was held from November 4th to 5th. This year's theme, "AI Together, AI Tomorrow," attracted more than 35,000 participants both online and offline, along with global AI leaders including OpenAI, Microsoft, and Nvidia. The event emphasized that close collaboration with partners is crucial for exploring the future of AI.
The reason experts commonly emphasize the importance of close collaboration is due to factors such as the lack of AI use cases, shortages of AI semiconductors, and energy issues. Despite ongoing large-scale investments in AI, there is still a lack of use cases and revenue models that can recover these investments. Additionally, large language models (LLMs) require an enormous amount of computation, leading many major tech companies to develop their own AI chips, but constraints still exist. Training the latest large language models requires tens of megawatts of power, and the LLMs expected to emerge next year are projected to demand hundreds of megawatts of power. As the growing significance of AI semiconductors, I realized the immense value of SK hynix Inc.'s efforts, which are focused on providing solutions in the semiconductor domain.
At this event, G-RTC also introduced computational memory, a technology to strengthen convergence technology capabilities, under the theme of “Embracing Memory Computing”. Emphasizing that research is the beginning of change, we presented the necessity of implementing an Open Research Platform and the value that the semiconductor ecosystem can obtain through ORP.
Through our participation in the SK AI Summit 2024, we were able to confirm that SK is well-prepared with a variety of strategies across the AI industry to the point where they can confidently align with the current trends. We also hope that SK hynix Inc. can leverage its global technological leadership to drive significant momentum for change, hitting new highs through close collaboration with its partners.

![[Summary] A Memristor-based In-Memory Computing SoC with Efficient Depthwise Convolution](https://mis-prod-koce-research-user-cdn-01-blob-ep.azureedge.net/web/blog/20260806/thumb_sEVmfv3K.20260806085519833.jpg)
![[Insights] Evolving Role of Emerging Memories in Next-Generation Computing](https://mis-prod-koce-research-user-cdn-01-blob-ep.azureedge.net/web/blog/20260806/thumb_NgJoNLYn.20260806080458980.jpg)
![[Summary] Electrical Characteristics of the 4F2 Vertical Gate (VG) DRAM integrated with Bit-Line Shielding (BLS) and Back Gate (BG) Transistor](https://mis-prod-koce-research-user-cdn-01-blob-ep.azureedge.net/web/blog/20260806/thumb_DOUurl5m.20260806080624234.png)