Podcast Summary: The Rundown
Episode: How Nebius is Becoming the AWS of AI (ft. Marc Boroditsky)
Host: Zaid Admani
Guest: Marc Boroditsky, Chief Revenue Officer at Nebius
Date: October 5, 2025
Overview
In this episode, host Zaid Admani interviews Marc Boroditsky, the Chief Revenue Officer of Nebius, one of the leading "Neo Cloud" companies rapidly rising amidst the AI infrastructure boom. Their discussion dives into Nebius’s unique positioning, its evolving relationships with industry giants like Nvidia and Microsoft, the intensifying demand for AI compute, and the physical and financial realities of data center expansion. The conversation pulls back the curtain on the "AWS of AI" narrative, competition, concerns about overbuilding, and what the next era of compute might look like.
Key Discussion Points and Insights
1. What Nebius Does & the “Neo Cloud” Category
- Full Stack AI Infrastructure:
- Nebius provides hardware, software, and services for AI engineers, helping them build, deploy, and scale AI applications seamlessly.
- Quote: “You can think of us...as the AWS of AI.” – Marc Boroditsky (01:19)
- Full Stack vs. Competitors:
- Differentiates from other players (like CoreWeave) by acting as a "NeoScaler": not just infrastructure, but an end-to-end platform for AI workloads.
- Quote: “The vision for the company is to meet the AI engineer where they are…They shouldn’t have to think about cobbling together or integrating or managing their environment.” – Marc Boroditsky (02:35)
2. Partnerships with Hyperscalers & The Microsoft Deal
- Major partnership with Microsoft: $17.4B+ agreement for Nebius to provide AI compute resources through 2031, including access to 100,000 Nvidia GB300 GPUs.
- Why Hyperscalers Partner Instead of Build:
- Explosive, unprecedented demand outpaces giants’ ability to build fast enough.
- Quote: “The preference is to build things out themselves. The pace of the market...is unprecedented.” – Marc Boroditsky (04:26)
- Nebius’s Role: Acts as a premium, flexible outsource and channel for hyperscalers.
3. Relationship with Nvidia
- Nebius has a "very special partnership" with Nvidia, including investment and collaborative programs (such as Inception for startups).
- Navigating Competition: Nvidia supplies several players but supports Nebius’s market flexibility.
- Quote: “Jensen’s taken very seriously the importance of relationships.” – Marc Boroditsky (05:52)
4. Why Nvidia Supports Neo Cloud Startups over Hyperscalers
- Hyperscalers (Amazon, Google) are developing their own chips, whereas Neo Clouds like Nebius are devoted Nvidia customers.
- Quote: “Our business is specialization. We are looking at the specific use cases and requirements that the AI engineer and the AI companies have.” – Marc Boroditsky (08:36)
5. Circular Financing & AI Bubble Concerns
- The AI infrastructure market is a multi-trillion-dollar shift, demanding aggressive capitalization and rapid capacity expansion.
- Addressed concerns about circular financing—where Nvidia invests in companies that, in turn, buy Nvidia hardware—as a necessary mechanism for growth at this scale.
- Quote: “Capitalization is critical ... Our approach is to make sure that we have the capital necessary to continue to grow at the rate that we are.” – Marc Boroditsky (11:14)
6. Physical Infrastructure & Power Challenges
- Building data centers requires huge investments, land, power, regulatory approvals, and logistics, not just chips.
- The biggest current constraint alternates between data center capacity and chip supply, but power is increasingly the limiting factor.
- Quote: “The pendulum swings...the most recent...half a year or so, it’s been data centers. We’re now getting into the main delivery period for the next generation of GPUs…” – Marc Boroditsky (12:59)
7. Evolution from Model Training to Inference
- Demand is rapidly shifting from simply training models (training) toward deploying them at scale (inference).
- Quote: “The technical term is inference...that’s where the request...generates a token and sends that back to the application...We are seeing an expansion in inference. That’s exciting to me.” – Marc Boroditsky (14:36)
8. On Overbuilding and Hardware Obsolescence
- Overbuilding is a recurring risk in tech waves, but fast fail cycles and enterprise adoption may mean resets and catch-ups happen faster.
- GPUs’ three-to-five-year lifespan is less of a risk while older chips can serve less demanding compute needs.
- Quote: “People are still using prior generation GPUs even though...the Blackwells and the GBs are just getting deployed. We’re selling the hell out of hoppers.” – Marc Boroditsky (20:48)
9. Energy Consumption and Regulation
- Surging electricity needs may push up local prices and trigger increased regulation, but Nebius can locate data centers flexibly—even in places like Iceland or Finland—to optimize costs and compliance.
- Quote: “We have a lot of flexibility…We’ve been building in places like Iceland and Finland and we as a supplier are looking for the right combination of location, capital and overall market opportunity.” – Marc Boroditsky (22:40)
10. The Future: Power Generation and Renewables
- Expectation that the AI boom could catalyze a new wave of renewable and nuclear energy projects, eventually lowering electricity costs long-term.
- Quote: “You’ve definitely seen a renaissance in power generation. There is no question about it.” – Marc Boroditsky (24:15)
Notable Quotes & Memorable Moments
- “We are the AWS of AI.” – Marc Boroditsky (01:19)
- On Nvidia: “Jensen’s taken very seriously the importance of relationships...He is a consummate statesman type CEO and very impressive.” – Marc Boroditsky (05:52, 28:24)
- On Overbuilding:
- “It’s happened in every single previous tech wave...The fail fast reality is going to make that reset period a lot faster.” – Marc Boroditsky (17:52)
- On Hardware Obsolescence:
- “I’ve actually got customers that have told me, ‘Can you get me 10,000 hoppers? I'll take any hoppers you have.’” – Marc Boroditsky (20:48)
- On AI Consumption:
- “We are seeing apps dramatically expand, actually.” – Marc Boroditsky (15:29)
- Jensen Huang’s Wardrobe:
- “He wore his black leather jacket the last time I saw him. He had to explain why he was wearing a suit...” – Marc Boroditsky (27:58)
- On AI Use in Business:
- “We’re in the middle of a GTM infrastructure buildout and we’re building an AI native GTM infrastructure.” – Marc Boroditsky (26:34)
- On Growth:
- “100%. 100%. That's the reality. I mean, if we could, we would.” – Marc Boroditsky (12:04)
Key Timestamps for Important Segments
- 00:57 – What Nebius actually does
- 02:12 – Differentiating Nebius from other Neo Cloud players
- 03:51 – Microsoft partnership details
- 05:07 – Microsoft’s access to Nvidia GPUs via Nebius
- 05:50 – Inside the Nebius-Nvidia relationship
- 07:50 – Why Nvidia favors Neo Clouds over hyperscalers
- 10:22 – Addressing AI bubble and circular financing concerns
- 12:58 – Physical constraints: data centers vs. chips
- 14:36 – Transition from model training to inference
- 17:29 – The risk and cycle of overbuilding
- 20:13 – Hardware lifespan concerns
- 22:29 – Regulatory risks and power challenges
- 24:15 – The power generation boom
- 25:18 – Reaction to new AI apps (OpenAI Sora)
- 26:30 – How Nebius uses AI internally
- 27:47 – Personal interactions with Jensen Huang
- 28:24 – Praise for Jensen Huang’s leadership style
Closing Thoughts
Marc Boroditsky offers a rare insider’s view into the intense growth and risk management happening within the booming AI infrastructure sector. Nebius positions itself as a flexible, full-stack alternative to both hyperscalers and single-layer offerings, with deep Nvidia partnerships fueling their ascendancy. The conversation touches not only on immediate business and technical challenges, but also the broader implications for power, regulation, and the pace of innovation.
Nebius’s rise hints at the birth of a new class of super-scalers, and the episode is essential for anyone looking to understand the next frontiers of AI infrastructure, market dynamics, and business model innovation in tech.
For Listeners
Missed the episode? This summary captures the core insights and big-picture trends discussed, giving you a detailed roadmap of Nebius’s approach and the emerging landscape of AI-powered cloud infrastructure.