Nebius $NBIS: Planning to expand to a 1GW-class data center in 2026 with the 'Neo Cloud' strategy, targeting the AI-native demand of emerging forces.
Arkady Volozh, CEO of Nebius $NBIS, appeared at DLD 25 held in Munich, Europe, revealing the company's plan to build a 1GW (Gigawatt) data center in 2026. This significantly exceeds the initial guidance (100MW in 2025, around 240MW in the mid-term), marking a noteworthy significant upward revision in the AI infrastructure market.
"There is readiness to enhance several hundred megawatts until next year, and to build infrastructure reaching 1GW by 2026 if the demand warrants it, further enhancement is possible."
The CEO emphasizes that in addition to the additional fundraising of over $0.7 billion from Nvidia $NVDA and VC giant Accel, backgrounded by "several billion dollars" in self-funding, a bold and aggressive expansion route was outlined.
Originally, a vast amount of CAPEX (capital investment) and advanced engineering are required for the construction and operation of 1GW class. In particular, new cloud providers must enter the market under different competitive conditions than the major players (such as AWS, Azure, Google Cloud). However, according to Mr. Arkady Volozh, Nebius has around 1,000 high-level talents mainly from the engineering division that handled Yandex's infrastructure, and has the experience of already building and operating data centers on a scale of several hundred megawatts.
"We are a team that has independently built the infrastructure of a major search engine in the past and was also one of Nvidia $NVDA's largest customers. Anticipating the new demand for AI infrastructure, we are well-versed in the optimal products and efficient server configurations."
He exudes confidence.
Funds raised from Nvidia $NVDA, Accel, and others completed by the end of the year support this strategy. Nebius announced a $0.7 billion capital increase and is considering further fundraising possibilities from the securities market as a listed company. As a global sales and development base, there are over 500 engineers based in Amsterdam, with R&D centers also established in Israel and the USA. Around 1000 top engineers from Yandex who left Russia have joined, consolidating expertise in cloud infrastructure and large-scale parallel computing.
"Although started as a European company, in reality, the AI infrastructure demand is significant in the USA. Therefore, the construction of several hundred megawatts will primarily proceed in the USA."
The CEO states.
Nebius's Target Materials are 'AI Natives,' startups focusing on generative AI (LLM) and large-scale data analysis, as well as independent software companies requiring unique model development.
Distinguishing itself from major clouds like AWS, Azure, and Google Cloud, Nebius aims to attract customers who compete with major tech companies as 'Neo Cloud.' Especially with considerations like privacy, cost structure, and technical uniqueness, there is a tailwind of companies increasing that do not want to rely on major clouds.
Furthermore, CEO Arkady Volozh stated, 'The actual demand is leading in the USA rather than Europe,' planning to rapidly build several hundred megawatts of datacenter in the USA first. Additionally, they will continue to expand their presence in Europe. Looking at the current customer portfolio, the high percentage of USA clients determines the company's investment priorities.
Furthermore, it's worth noting that Nebius $NBIS is not just a cloud vendor but also pursuing a diversification strategy.
Taloa: A data labeling business that handles the creation of training data for LLM and AI systems. Its main clients are companies building large models like Google and Microsoft.
Reskilling (boot camp) business: Providing short-term intensive technical training for working adults looking to transition into the AI/data field. This business is expected to boost the AI talent market where skills shortage has become critical.
Autonomous Driving: A team with a proven track record of public road unmanned driving has joined, foreseeing 2025 as a 'turning point for mainstream adoption.' Due to the extensive driving data required, it has significant synergy with AI infrastructure.
By these multiple businesses mutually complementing each other, Nebius has the potential to build an 'end-to-end AI ecosystem.' Being able to complete a series of flows internally, from AI model training (Taloa) to human resource training (Reskilling) to advanced applications like autonomous driving, allows investors to see significant long-term growth potential as a major attraction.
📍Future Prospects
Led by Mr. Arkady Volozh, Nebius (NBIS) leverages the expertise in operating large-scale search infrastructure nurtured at Yandex, a strong partnership with Nvidia (NVDA), and a capital strength of tens of billions of dollars as weapons, aiming for a 1GW expansion as 'Neo Cloud' by 2026. This plan significantly exceeds existing guidance and has attracted strong interest in both risk and return perspectives in the market. If the company can establish recognition as a new option to challenge major clouds by successfully building AI infrastructure as planned, there could be significant rewards for investors.
However, implementing a 1GW-class data center in a short period requires a massive investment in equipment and engineering resources, and if the plan fails, the risk of losses will materialize. On the other hand, as the demand for AI infrastructure continues to increase, sectors requiring high-speed processing and large capacity are expected to expand steadily. Nebius' overwhelming pace of facility expansion can be seen as a kind of pioneering strategy aimed at the 'supply and demand tightness' in the AI boom.
In fact, top-tier AI startups in the USA are rapidly increasing cloud demand. In cases where mega-clouds cannot allocate sufficient resources or face cost-related challenges, there is significant room for welcoming 'alternative infrastructures' like Nebius.
The need for 'avoiding dependence on competitors' among AI-native companies and demands for data sovereignty in Europe and emerging countries are additional tailwinds, making aggressive CAPEX investments likely to pay off.
Right now, in emerging fields with high demand such as generative AI and self-driving cars, a new force that can provide substantial processing power is emerging. This phase, which can be called the 'next chapter' of the AI infrastructure market, will see investors carefully watching for the day Nebius, playing a part in that chapter, actually puts the 1GW-class into operation.
"There is readiness to enhance several hundred megawatts until next year, and to build infrastructure reaching 1GW by 2026 if the demand warrants it, further enhancement is possible."
The CEO emphasizes that in addition to the additional fundraising of over $0.7 billion from Nvidia $NVDA and VC giant Accel, backgrounded by "several billion dollars" in self-funding, a bold and aggressive expansion route was outlined.
Originally, a vast amount of CAPEX (capital investment) and advanced engineering are required for the construction and operation of 1GW class. In particular, new cloud providers must enter the market under different competitive conditions than the major players (such as AWS, Azure, Google Cloud). However, according to Mr. Arkady Volozh, Nebius has around 1,000 high-level talents mainly from the engineering division that handled Yandex's infrastructure, and has the experience of already building and operating data centers on a scale of several hundred megawatts.
"We are a team that has independently built the infrastructure of a major search engine in the past and was also one of Nvidia $NVDA's largest customers. Anticipating the new demand for AI infrastructure, we are well-versed in the optimal products and efficient server configurations."
He exudes confidence.
Funds raised from Nvidia $NVDA, Accel, and others completed by the end of the year support this strategy. Nebius announced a $0.7 billion capital increase and is considering further fundraising possibilities from the securities market as a listed company. As a global sales and development base, there are over 500 engineers based in Amsterdam, with R&D centers also established in Israel and the USA. Around 1000 top engineers from Yandex who left Russia have joined, consolidating expertise in cloud infrastructure and large-scale parallel computing.
"Although started as a European company, in reality, the AI infrastructure demand is significant in the USA. Therefore, the construction of several hundred megawatts will primarily proceed in the USA."
The CEO states.
Nebius's Target Materials are 'AI Natives,' startups focusing on generative AI (LLM) and large-scale data analysis, as well as independent software companies requiring unique model development.
Distinguishing itself from major clouds like AWS, Azure, and Google Cloud, Nebius aims to attract customers who compete with major tech companies as 'Neo Cloud.' Especially with considerations like privacy, cost structure, and technical uniqueness, there is a tailwind of companies increasing that do not want to rely on major clouds.
Furthermore, CEO Arkady Volozh stated, 'The actual demand is leading in the USA rather than Europe,' planning to rapidly build several hundred megawatts of datacenter in the USA first. Additionally, they will continue to expand their presence in Europe. Looking at the current customer portfolio, the high percentage of USA clients determines the company's investment priorities.
Furthermore, it's worth noting that Nebius $NBIS is not just a cloud vendor but also pursuing a diversification strategy.
Taloa: A data labeling business that handles the creation of training data for LLM and AI systems. Its main clients are companies building large models like Google and Microsoft.
Reskilling (boot camp) business: Providing short-term intensive technical training for working adults looking to transition into the AI/data field. This business is expected to boost the AI talent market where skills shortage has become critical.
Autonomous Driving: A team with a proven track record of public road unmanned driving has joined, foreseeing 2025 as a 'turning point for mainstream adoption.' Due to the extensive driving data required, it has significant synergy with AI infrastructure.
By these multiple businesses mutually complementing each other, Nebius has the potential to build an 'end-to-end AI ecosystem.' Being able to complete a series of flows internally, from AI model training (Taloa) to human resource training (Reskilling) to advanced applications like autonomous driving, allows investors to see significant long-term growth potential as a major attraction.
📍Future Prospects
Led by Mr. Arkady Volozh, Nebius (NBIS) leverages the expertise in operating large-scale search infrastructure nurtured at Yandex, a strong partnership with Nvidia (NVDA), and a capital strength of tens of billions of dollars as weapons, aiming for a 1GW expansion as 'Neo Cloud' by 2026. This plan significantly exceeds existing guidance and has attracted strong interest in both risk and return perspectives in the market. If the company can establish recognition as a new option to challenge major clouds by successfully building AI infrastructure as planned, there could be significant rewards for investors.
However, implementing a 1GW-class data center in a short period requires a massive investment in equipment and engineering resources, and if the plan fails, the risk of losses will materialize. On the other hand, as the demand for AI infrastructure continues to increase, sectors requiring high-speed processing and large capacity are expected to expand steadily. Nebius' overwhelming pace of facility expansion can be seen as a kind of pioneering strategy aimed at the 'supply and demand tightness' in the AI boom.
In fact, top-tier AI startups in the USA are rapidly increasing cloud demand. In cases where mega-clouds cannot allocate sufficient resources or face cost-related challenges, there is significant room for welcoming 'alternative infrastructures' like Nebius.
The need for 'avoiding dependence on competitors' among AI-native companies and demands for data sovereignty in Europe and emerging countries are additional tailwinds, making aggressive CAPEX investments likely to pay off.
Right now, in emerging fields with high demand such as generative AI and self-driving cars, a new force that can provide substantial processing power is emerging. This phase, which can be called the 'next chapter' of the AI infrastructure market, will see investors carefully watching for the day Nebius, playing a part in that chapter, actually puts the 1GW-class into operation.
Disclaimer: Community is offered by Moomoo Technologies Inc. and is for educational purposes only.
Read more
Comment
Sign in to post a comment
Market Insights
Hot Topics Hot Topics
Trump's 'Taco deal' sparks market rebound : Market rebound sustainable ?
1. If tariffs pause, which sectors benefit first? 2. Trump's "threat-compromise" cycle normalizes—how to hedge across assets?
Moo Options Explorer
Apr 7 17:28
Weekly roundup: Tariff turmoil and recession fears dominate
Moo Options Explorer
Mar 31 17:01
Weekly roundup: Tariff concerns and Tesla’s rollercoaster ride
めいちゃんの羊 : Nebius, NVIDIA.