Nebius Group N.V. (NBIS) Earnings Call Transcript & Summary
October 18, 2024
Earnings Call Speaker Segments
John Boynton
executiveHello. I'm John Boynton, Chairman of Nebius, and I'm here presenting with our CEO, Arkady Volozh. On Monday, our shares will start trading again on the NASDAQ. The road to this moment has been long and complex, but we are very excited to have reached this new chapter in our development. Resumption of trading is the right catalyst for us to tell you more about who we are and the tremendous market opportunity that lies ahead of us. We are at the very beginning of the AI revolution. Nobody can be sure which business models or underlying technologies will prevail, but we can be sure of one thing. The demand for AI infrastructure will be massive and sustained. This is the market space where Nebius will play. Let me put Nebius in some historical context. A number of us actually worked together at Yandex for over 20 years, building what became one of Europe's largest technology companies. After Russia's invasion of Ukraine in 2022, our worlds changed. Half our senior management and hundreds of our top engineers left Russia for good. It was at that point that we set out to start building something new. That something is Nebius. While we managed our separation from Russia, our R&D team was busy pivoting to new business models, developing original technologies in code and finding fresh opportunities for growth. Yandex was best known as a search engine and consumer Internet brand and over the years, the business expanded considerably. Our R&D team gained substantial experience in building large-scale data centers, cloud, computing architecture and other digital infrastructure. Their technological expertise puts us in very good stead now as we set out to build one of the world's largest AI infrastructure companies.
Arkady Volozh
executiveYes, it's in the AI infrastructure space, where we see a huge opportunity. Building effective, optimized and large-scale AI infrastructure is a challenge for the whole industry. It requires unique and specific expertise, which is what we have. It's not just about feeding all data centers and plugging in new chips. Everything has to be built from the ground up and specifically tailored from the outset to meet the needs of AI. It's hard to believe that neural networks were declared kind of a voodoo technology in the mid-2000s. This was until we all realized that it's just because of not enough compute. Compute is what underpins all this current revolution in GenAI, and increasing demand for compute will fuel the growth of this industry. We have seen substantial growth over the past few years, and we believe it's still just early days. Our ambition is to build data centers with hundreds of megawatts providing AI compute infrastructure and value-added services to the global AI industry. Let me give you an overview of our business and growth perspectives. We have our main headquarters and largest R&D presence in Amsterdam, where we have around 250 people now. We have additional R&D hubs in other parts of Europe as well as in Israel. And we are rapidly expanding our presence in the U.S., which is where many of our clients are. Simply put, our business is based on two critical components: technology and capital. Although we are a relatively new company, we already possess both. As publicly disclosed, we have approximately $2 billion in cash, hundreds of top AI engineers and state-of-the-art data center infrastructure. We believe that this initial capital, combined with our talent and our strong relationship with NVIDIA, positions us to start scaling quickly and significantly boost our compute capacity. Moreover, as a public company, we will benefit from access to a wider range of capital sources, which would enable us to go even faster. We believe that given the magnitude of required compute and the associated costs, the market will be dominated by big players. These are obviously the hyperscalers, Google, Amazon, Microsoft, large private clouds like Meta, Apple, Tesla, and just a few additional sizable specialized players. This is where we fit in. The hyperscalers and the large private clouds will likely consume most of their compute internally, leaving most of the independent market to a handful of specialized players like us, with capacity of hundreds of thousands of GPUs. Why we are so confident in our ability to play such a significant role? Our expense is proven. We have expertise in running data centers with hundreds of megawatts power loads and building world-class cloud and services. Let me be perfectly clear. We've done this before, and we will do it again. It's important to remember that compute power is critical for our AI clients. They expect reliability, flexibility, fast adaptations and first-class technical support. We meet and exceed the expectations. We believe that focus and specialization are essential. From R&D, hardware and software into technical support and pricing, this is exactly where we're concentrating our efforts. This is exactly where we excel. We design and build our data centers and our server racks in-house, leveraging our unique technical expertise and know-how. This enables us to operate in more efficiently and with greater reliability. In turn, this leads to higher margins or better price affordability for clients, enabling us to boost our market share. We do much more than that. On top of the hardware, we have built an AI native cloud computing environment, specifically for AI workloads. Our full stack infrastructure includes compute, storage, managed services and additional tools. So it is a much wider product portfolio than just a GPU Cloud. We build a fully virtualized infrastructure, a complete development environment for AI developers, all assembled through a single intuitive platform. On the data center front, as already announced, we are expanding our data centers in Finland to 75 megawatts of compute power. Additionally, we have leased a data center in Paris. And we are in the process of further expanding our data center presence both in Europe and in the U.S. and expect to exceed hundreds of megawatts of capacity in the short to medium term. With respect to the demand, as we all know, there is currently massive global demand for AI compute power. And this will continue to grow, we believe. We're also seeing significant capital investments being made into AI companies building their own models or developing their new AI applications. We believe that a large portion of that capital will be directed toward compute infrastructure. And we expect this demand for compute to continue growing sustainably in the coming years. The company's that we'll be able to provide large-scale, efficient and reliable compute, will be in the best position to capture a major share of this market. We aim to be one of those key players. We plan to direct most of our CapEx to purchase NVIDIA Blackwell GPUs. We believe the demand of this -- for these GPUs will remain strong for the years to come. We have already secured a substantial number of GB200s with NVIDIA for 2025, next year. [indiscernible] the approximately 20,000 of H100s and H200s, which we will be deployed by the end of this year. We also anticipate continued demand for our existing H100s and H200s in the coming years insurance will play an increasingly important role in the future. And with our ability to offer compute power via tokens, we expect clients to continue utilizing our compute services for the long term. Nebius is our core business, and that's the one we are mostly focusing on today. But we do also own a number of other AI-related businesses, each of which has a great production. For example, Toloka. Toloka offers solutions that provide high-quality expert data at scale for the GenAI industry. Toloka is one of only a few companies globally that is set up to do this. That's along side with the likes of Scale AI and Surge AI, for example. Tripleten. Tripleten Is a leading ad tech platform that focuses on reskilling people for successful careers in tech. The team primarily focuses on the U.S. and Latin American markets and was named the best overall software engineering [indiscernible] in the U.S. by Fortune magazine in June 2024 this year. Avride. Avride is uniquely positioned in the autonomous vehicle industry by simultaneously developing autonomous driving technology for self-driving cars and deliver [indiscernible] The team has recently signed a multiyear strategic partnership with Uber for both mobility and delivery and is actively preparing to launch [indiscernible] in Austin in the coming weeks. We will be flexible in our strategic approach to these businesses, and we look to bring in strategic partners to investors.
John Boynton
executiveThank you, Arkady. And by the way, Arkady and I have been partners for 34 years. Okay. Let me recap. Here's my take on what makes Nebius such an exciting business. First, we have our core AI infrastructure business, which is strongly positioned to succeed for several reasons. We are a pure-play AI company in its most fundamental segment, AI infrastructure. Wherever AI goes, it will always need infrastructure. Our full stack spans the AI infrastructure value chain from data centers and in-house design servers, to AI cloud platforms and managed services for the AI world. We have a long-standing collaboration with NVIDIA, giving us ongoing access to the latest and fastest GPUs. And most important, we have an incredible team with the right experience, skills and track record to drive our growth. As you may know, our divestment left us with a cash balance of around $2 billion, and we have good access to capital markets. We have three additional businesses, each of them among the leading and most promising in their respective sectors. And finally, we have significant minority economic interest in ClickHouse, which is a business that we founded, developed in-house and successfully spun off in 2021 and which was valued at $2 billion in its second financing round later that year. The road to today was not an easy one. But I could not be more excited about the road we have chosen. The opportunity ahead of us is immense, and we have the right technology and team for this moment. And we have the ambition to build something even bigger than what we built before. With trading now set to resume, we look forward to fully reengaging with the market as we embark on this very exciting new chapter. Thank you.
This call discussed
For developers and AI pipelines
Programmatic access to Nebius Group N.V. earnings transcripts and 251,000+ others is available through the
EarningsCalls.dev REST API. Plans from $24.99/month — full transcripts, speaker segments,
full-text search, and the recently-added /api/v1/transcripts/recent polling endpoint for ETL pipelines.