[Seminar] NVIDIA's Changed Competitive Strategy for AI Data Centers
[Lecturer]
Yukio Sakaguchi, Representative of the Clean Energy Research Institute
[Key Lecture Content]
This lecture will organize what has fundamentally changed in AI data centers, starting from the evolution of NVIDIA. Currently, NVIDIA's competition is rapidly shifting from GPU performance itself to the supporting factors of power, cooling, water, and location conditions. Particularly due to advancements in high density, traditional air cooling-centered designs are becoming difficult to manage, necessitating new design philosophies that include liquid cooling and high-temperature water cooling.
First, based on NVIDIA's roadmap and changes in rack design, we will clarify why the constraints of AI data centers have shifted from semiconductors to infrastructure. Next, we will outline the characteristics and limitations of air cooling, liquid cooling, and high-temperature water cooling, and consider how external factors such as water constraints, power infrastructure, grid connection, and location conditions impact cooling methods and business viability. Furthermore, taking into account changes in infrastructure investment, power procurement, and risk sharing, we will indicate where realistic business opportunities may lie in the future. This will not only focus on technical discussions but also present a perspective that reconsiders AI infrastructure as a condition for business viability.

| Date and time | Thursday, Jun 18, 2026 02:00 PM ~ 05:00 PM |
|---|---|
| Capital | Venue attendance / Live streaming / Archive streaming (available for viewing multiple times for 2 weeks) SSK Seminar Room, 2-6-2 Nishi-Shinbashi, Minato-ku, Tokyo, 4th Floor, Zaimax Nishi-Shinbashi Building |
| Entry fee | Charge |
More Details & Registration
Details & Registration



