Sparks of Insight: Bridging Power and Compute with Jimmy Hayes
- Rich Washburn
- Jun 30
- 2 min read
I always walk away from conversations with new perspectives, and this morning’s chat with Jimmy Hayes at Data Power Supply was no exception. Jimmy has a rare, 360° view of data-center deployments—everything from reciprocating engines and microgrids to UPS systems and switchgear.
In most projects, power and compute teams operate in parallel universes: one is sizing transformers while the other is racking NVIDIA H100 GPUs. Rarely do they collide—and that’s where crucial gaps emerge. Jimmy’s mastery lies in uniting those worlds. Here are three takeaways that sparked fresh thinking for me:
1. Ink, Press, and Paper for AI
Think of your data center as a printing press. The compute (GPUs, servers) is the press itself, but power is your ink and paper. Buy the latest AI accelerators, and without the right electrical infrastructure, they’ll sit idle—expensive bricks. Jimmy’s team delivers the “ink”—modular power pods, microgrids, and rapid-deploy switchboards—so clients hit the ground running.
2. Hash Power as the Modern Arsenal
Whether you’re mining Bitcoin or training massive language models, the thirst for gigawatts is the same. While hyperscalers explore nuclear solutions, mid-tier players need agility. Jimmy’s approach? Renewable-friendly, scalable power solutions that grow alongside your compute demands—no six-month grid backlogs required.
3. Utility Relationships Unlock Speed
Securing a new substation upgrade can feel like waiting for rain in a drought. Data Power Supply’s CTO, Jeff Robinson, brings 30+ years of utility partnerships to the table—backstage passes to the grid that turn month-long wait times into a matter of weeks.
Big thanks to Jimmy Hayes for sharing these insights. If your build isn’t integrating power and compute teams from day one, you’re only telling half the story. Conversations like this remind me: true innovation happens when disciplines collide. Here’s to more sparks of insight—and to powering the AI revolution end-to-end.
留言