From Grid to Chip: Designing the Future of AI Data Centers
The next wave of AI will be decided not just by algorithms, but by where compute lives and how we power it. Data centers are becoming the new cities of intelligence — concentrated, capital-intensive, and foundational to economic and technological abundance. Yet as demand for AI-scale compute surges, the biggest constraint isn’t silicon — it’s the grid.
This panel brings together leaders who are building that future, interrogating market signals, hyperscaler intent, and the practical tradeoffs that define a future-ready site:
Yubo She, Technical Accounting @ OpenAI
Satej Desai, Data Center Engineering @ NVIDIA
Phil Burt, Client Executive @ STACK Infrastructure
Moderated by: Shaneez Mohinani, Strategy & Operations @ GridCARE
The panel will share perspectives on:
Market & hyperscaler sentiment: How are hyperscalers thinking about growth, risk, and site strategy right now? Are they concentrating in mega hubs or diversifying geographically — and what signals should the market be reading?
The future we’re building toward: In 5–10 years, what does the compute landscape look like — centralized megasites, distributed edge, or a hybrid? What operational and architectural shifts are inevitable?
Scale of investment: What is the scale and shape of capital required to deliver that future?
Overcoming the power bottleneck: How can the industry accelerate power availability for AI workloads without compromising grid reliability or affordability? What new models of collaboration between utilities, developers, and AI operators will define the next decade?
GridCARE is hiring — if you're excited about powering the AI era, we'd love to meet you!