Who’s Cutting the Cost of AI? — Crusoe × Nebius | AI Infra Summit
Who’s Cutting the Cost of AI?
AI is getting more powerful — but running it at scale is still expensive.
At this year’s AI Infra Summit, join us for a discussion on where AI cost is really created — from power and data center capacity, to cloud infrastructure, inference, and the software layer.
Rather than looking at one piece of the stack, we’ll ask a bigger question:
What is actually bringing down the cost of AI — and where is there still the most waste?
Key Discussion Topics
Power, Capacity & Deployment Speed: How faster modular infrastructure and more flexible capacity planning can reduce last-minute premiums and infrastructure risk.
Vertical Integration: How connecting data centers, cloud, compute, and inference can remove inefficiencies across the AI stack.
Inference & Software Efficiency: How better utilization, model optimization, serving, and software can generate more useful output from the same underlying compute.
The End-User Perspective: What large AI users actually care about when buying compute, and which improvements genuinely change their economics.
Featured Speakers
Ryan Hanrui Wang — SVP of Applied AI at Nebius | Co-Founder & CEO at Eigen AI
Michael Yen-Chi Ho — Director of Technical Product Marketing at Crusoe
Final End-User Speaker — To Be Announced
Moderator: Yuri Cheng — Partner at Yotta Core Energy
*** Please note that conference ticket/access is required to attend this panel***
Join Us at AI Infra Summit
Wednesday, September 16
3:30 PM – 4:15 PM
Santa Clara Convention Center
Use code WOMENINAICLUB15 at registration for 15% off your ticket.
-----
🏛️ About the Host
Women in AI is a community supporting female founders, investors, engineers, and leaders shaping the future of AI.
💜 Become a member for priority access to future events
💜 Follow us on LinkedIn, X, and Instagram
📩 Join our newsletter
💜 Interested in sponsoring or partnering? Book time here
Yotta Compute Energy (YCE) delivers high-performance, enterprise-grade GPU cloud infrastructure engineered for compute-intensive AI training, fine-tuning, and real-time inference. It provides model builders, startups, and enterprises with the reliable, cost-efficient compute required to scale mission-critical AI workloads seamlessly.