Building Cloud for AI Workloads

What this episode covers

Building a cloud for AI is more than just stacking GPUs together—it requires deep software expertise and hardware innovation working in sync. At NVIDIA GTC, I had the chance to sit down with Gleb K. and Oleg Fedorov from Nebius to break down what it really takes to build a high-performance AI cloud from scratch.Gleb, leading the software side, shared insights into the complexities of AI cloud infrastructure, while Oleg, focusing on hardware, talked about the compute power needed to support today’s AI demands.We covered:✅ What it means to be a "serial cloud builder" and how Gleb transitioned from managing data centers to architecting AI-first cloud solutions✅ Why an AI cloud is much more than just a collection of GPUs and how Nebius optimized for speed and efficiency✅ The key ingredients of “good compute” and what it takes to deliver top-tier performance in AI workloads✅ How Nebius and NVIDIA are working together to push the boundaries of AI infrastructureThis interview is about the software-hardware synergy that powers AI innovation and the future of AI cloud computing.#data #ai #nebius #nvidiagtc2025 #theravitshow

PodParley-generated summary based on available episode metadata and transcript content.